JP7689015B2 - Waste quality detection system - Google Patents

Description

The present invention relates to a garbage quality detection system.

The collected waste contains low-quality waste with low combustion efficiency, such as waste with a high moisture content. If the proportion of low-quality waste in the waste put into the incineration facility fluctuates significantly, stable combustion becomes difficult. For this reason, the collected waste is temporarily stored in a pit and mixed using a crane. After the waste has been homogenized by mixing, it is put into the incineration facility.

To homogenize the waste efficiently, it is important to be able to determine the quality of the waste for each specified area of the waste. For example, the waste can be homogenized efficiently by using a crane to grab the waste from an area determined to have a high content of low-quality waste and moving it to an area determined to have a low content of low-quality waste. Technology for determining the quality of the waste in a pit is disclosed in JP 2016-216228 A and JP 2019-160153 A.

In the pit crane control system described in JP 2016-216228 A, an infrared irradiation device irradiates infrared rays toward the garbage, and an imaging device captures the reflected light. In this image, the higher the brightness of the area, the less moisture there is, and the garbage quality is determined.

In the garbage quality estimation system of JP2019-160153A, the image captured by the imaging unit is corrected based on height information measured by a laser rangefinder so that all areas of the garbage are on the same height plane. This image is divided into evaluation areas, and the variation in the area of light or dark areas in each area is calculated to evaluate the degree of mixing.

JP 2016-216228 A JP2019-160153A

The height of the garbage stored in the pit is not constant. The surface of the garbage is uneven. Meanwhile, the image data of the garbage surface photographed to determine the garbage quality is a two-dimensional image. It is possible that the area of the image data where the garbage quality has been determined will not correspond to the actual area on the surface of the garbage. In this case, for example, even if a crane bucket is used to grab an area where the garbage quality has been determined to be poor from the image data, it will end up grabbing garbage that is out of alignment with this area. This can be an obstacle to efficiently standardizing the garbage quality.

The object of the present invention is to provide a litter quality detection system that can accurately match the area in the litter image data from which the litter quality is detected with the area on the actual litter surface.

The present invention relates to a garbage quality detection system for garbage stored in a pit. This garbage quality detection system includes a photographing device that photographs the garbage to obtain image data, a level gauge that measures the height of the garbage at a specified horizontal position, a boundary calculation device that calculates the boundary in the image data of a specified virtual area that serves as a unit area when evaluating garbage quality based on the horizontal position and height and the image data, and a feature calculation device that finds the feature values of the garbage in the virtual area from the image data within the boundary.

Preferably, the boundary calculation device calculates the boundary of the virtual area in the image data based on the horizontal positions and heights of the vertices of the virtual area on the surface of the garbage.

The boundary calculation device may calculate the boundary of the virtual area in the image data based on the horizontal position and height of the vertex of the virtual area at the average height position of the garbage in the virtual area.

Preferably, the waste quality detection system further includes a first calibration device. The first calibration device performs calibration to associate coordinates in the image data with positions in the pit based on image data captured by the imaging device by positioning a specific measurement point of a crane used in the pit at multiple different locations in the pit.

Preferably, the litter quality detection system further includes a second calibration device. The second calibration device performs calibration to correlate the measurement results of the level meter with positions in the pit based on the horizontal coordinate and height of a predetermined measurement point on a first side wall of the pit and the horizontal coordinate and height of a predetermined measurement point on a second side wall perpendicular to the first side wall, both measured by the level meter.

Preferably, the characteristic amount is the amount of moisture contained in the garbage.

The present invention relates to a method for detecting the quality of garbage stored in a pit, the method comprising the steps of:
A step of photographing the surface of the rubbish with an imaging device to obtain image data;
measuring the height of the waste at a predetermined horizontal position with a level meter;
The method includes the steps of: calculating the boundary in the image data of a specified virtual area that serves as a unit area for evaluating litter quality based on the horizontal position and height and the image data; and determining the characteristic quantities of litter in this virtual area from the image data within the boundary.

Preferably, the litter quality detection method further includes a step of calibrating the image capturing device, the step of calibrating the image capturing device comprising:
The method includes the steps of: positioning specific measurement points of the crane at a number of different locations within the pit, photographing them using the photographing device, and obtaining image data; and performing a calibration step to associate the coordinates in the image data of the measurement points with their positions in the pit, based on the coordinates in the image data of the measurement points.

Preferably, the method for detecting litter quality further includes a step of calibrating the level meter. The step of calibrating the level meter includes:
The method includes the steps of: measuring, using the level meter, the horizontal coordinate and height of a specified measurement point on a first side wall of the pit, and the horizontal coordinate and height of a specified measurement point on a second side wall perpendicular to the first side wall; and performing calibration based on the results of the measurements to correlate the measurement results of the level meter with a position in the pit.

The waste storage facility of the present invention comprises a pit for storing waste, a crane for stirring the waste, and a waste quality detection system. The waste quality detection system comprises a photographing device for photographing the waste and obtaining image data, a level gauge for measuring the height of the waste at a predetermined horizontal position, a boundary calculation device for calculating the boundary in the image data of a predetermined virtual area that serves as a unit area when evaluating waste quality based on the horizontal position and height and the image data, and a feature calculation device for determining the feature values of the waste in this virtual area from the image data within the boundary. The crane stirs up the waste based on the feature values of the waste.

In the litter quality detection system of the present invention, the boundaries of virtual areas in the image data are calculated from image data of the litter captured by the camera device and the height of the litter at a specified horizontal position measured by the level meter. The litter features are calculated from the image data within the boundaries of each virtual area. This allows the area in the image data where the litter features have been calculated to correspond accurately to the area of the actual litter surface. By using this litter quality detection system, efficient uniformity of litter quality can be achieved.

FIG. 1 is a schematic diagram of a waste storage facility including a waste quality detection system according to one embodiment of the present invention. FIG. 2 is an enlarged schematic view of a portion of the installation of FIG. FIG. 3 is a configuration diagram of a waste quality detection system according to an embodiment of the present invention. FIG. 4 is a schematic diagram showing the calibration of the photographing device of FIG. FIG. 5 is a schematic diagram showing the calibration of the level meter of FIG. FIG. 6(a) is an example of a boundary of a virtual area in image data of garbage calculated by the garbage quality detection system of FIG. 3, and FIG. 6(b) is a diagram showing an example of boundary points for calculating this boundary. FIG. 7 is a block diagram showing a litter quality detection system according to another embodiment of the present invention. FIG. 8(a) is an example of the boundary of a virtual area in image data of garbage calculated by the garbage quality detection system of FIG. 7, and FIG. 8(b) is a diagram showing an example of boundary points for calculating this boundary.

The present invention will now be described in detail based on a preferred embodiment, with reference to the drawings as appropriate.

[First embodiment]
Fig. 1 is a schematic diagram showing a waste storage facility 1 equipped with a waste quality detection system 2 according to the present invention. The waste storage facility 1 includes a pit 4, a crane 6, and the waste quality detection system 2. Fig. 1 also shows a hopper 8, which is an inlet for waste to an incinerator (not shown), and waste (stored waste 10) stored in the pit 4. Fig. 2 is a schematic diagram showing an enlarged view of the pit 4, a portion of the crane 6, and the waste quality detection system 2.

The pit 4 temporarily stores garbage. The pit 4 is box-shaped and surrounded by four side walls 12 (first side wall 12a, second side wall 12b, third side wall 12c, and fourth side wall 12d) and a floor. The normal direction of the first side wall 12a is perpendicular to the normal direction of the second side wall 12b and the normal direction of the fourth side wall 12d. The normal direction of the third side wall 12c is perpendicular to the normal direction of the second side wall 12b and the normal direction of the fourth side wall 12d. The normal direction of the floor is perpendicular to all of the normal directions of these four side walls 12. Although not shown, at least one side wall 12 is provided with an input door for inputting garbage into the pit 4.

In order to identify a position within the pit 4, three-dimensional reference coordinate axes are defined in the pit 4. In FIG. 2, these coordinate axes are shown as x-axis, y-axis, and z-axis, which are mutually orthogonal. The x-axis and y-axis are horizontal coordinate axes, and the z-axis is vertical coordinate axis. In this embodiment, the x-axis is parallel to the normal direction of the first side wall 12a and the normal direction of the third side wall 12c of the pit 4, and the y-axis is parallel to the normal direction of the second side wall 12b and the normal direction of the fourth side wall 12d. The z-axis is parallel to the normal direction of the floor. The reference coordinate axes are not limited to this method of determination. It is sufficient that the reference coordinate axes are determined so that a three-dimensional position within the pit 4 can be identified.

In the pit 4, a "virtual area 14" is defined as a unit area for evaluating the quality of the garbage. In FIG. 2, an example of the virtual area 14 is shown by a two-dot chain line. As shown in FIG. 2, a plurality of virtual areas 14 are arranged two-dimensionally in a plan view. The pit 4 is divided into a plurality of virtual areas 14. In this embodiment, the pit 4 is divided into virtual areas 14 by a plurality of virtual division lines extending parallel to the x-axis and a plurality of virtual division lines extending parallel to the y-axis. Each virtual area 14 is rectangular in a plan view. In this specification, a point having the same horizontal coordinate as a vertex of this rectangle in a plan view is referred to as a vertex of this virtual area 14. The position and size of each virtual area 14 are predetermined. As will be described later, the virtual area 14 is used for positioning the crane 6 to grab garbage when the crane 6 is used to agitate the garbage. In this embodiment, the size of the virtual area 14 is set to a size that the crane 6 can grab at one time.

1, the crane 6 includes a pair of rails 16 extending in the front-rear direction (perpendicular to the plane of the drawing in FIG. 1), a girder 18 suspended between the rails 16, a trolley 20 positioned on the girder, a rope 22 extending from the trolley 20, a bucket 24 suspended from the trolley 20 via the rope 22, and a drive controller 26. The bucket 24 can move forward, backward, left, and right as the girder 18 moves forward and backward on the rails 16 and the trolley 20 moves left and right on the girder 18. The bucket 24 can move up and down as the trolley 20 winds up or down the rope 22. The bucket 24 can move to a desired position in the pit 4. The bucket 24 can be opened and closed. The bucket 24 can grab or release garbage. The drive controller 26 controls the movement of the girder 18 and the trolley 20 to move the bucket 24 to a desired position. The drive controller 26 can control the opening and closing of the bucket 24. This allows the crane 6 to grab the stored waste 10 at a given location and move it to another given location. The crane 6 can grab waste with the bucket 24 and dump the waste into the hopper 8.

The drive controller 26 is capable of grasping the position of the girder 18, the position of the trolley 20, and the length of the rope 22 on the reference coordinate axis. The drive controller 26 is capable of grasping the coordinate of the bucket 24 on the reference coordinate axis. The drive controller 26 can move the bucket 24 to a specified coordinate on the reference coordinate axis.

The structure of the crane 6 is not limited to the above structure. The crane 6 only needs to be able to grab the stored waste 10 at a specified position in the pit 4 and move it to the specified position.

The garbage quality detection system 2 detects the garbage quality of the stored garbage 10 in the pit 4. The detected garbage quality is sent to the drive controller 26 of the crane 6 and is used to agitate the stored garbage 10 by the crane 6.

Figure 3 shows the configuration of a litter quality detection system 2 according to one embodiment of the present invention. The litter quality detection system 2 includes an imaging device 28, a level meter 30, a first calibration device 32, a second calibration device 34, a boundary calculation device 36, and a feature calculation device 38. This figure also shows the internal configuration of the boundary calculation device 36, which will be described later. In this embodiment, the first calibration device 32, the second calibration device 34, the boundary calculation device 36, and the feature calculation device 38 are realized by a calculator (computer), a memory, and a program. These are realized by operating the calculator with a program. In Figure 2, the first calibration device 32, the second calibration device 34, the boundary calculation device 36, and the feature calculation device 38 are represented by a single box.

The photographing device 28 photographs the stored garbage 10 in the pit 4 to obtain image data. As shown in FIG. 2, the photographing device 28 is attached to the upper end of the side wall 12. The photographing device 28 is attached at a position where all virtual areas 14 can be photographed. In this embodiment, the photographing device 28 is attached to the upper end of the third side wall 12c. The position where the photographing device 28 is attached is not limited to the third side wall 12c. The photographing device 28 may be attached to another position as long as all virtual areas 14 can be photographed. In this embodiment, there is one photographing device 28. A plurality of photographing devices 28 may be attached to the pit 4. For example, two photographing devices 28 may be attached to the pit 4, one photographing device 28 photographs half of the virtual areas 14, and the other photographing device 28 photographs the remaining half of the virtual areas 14.

Although not shown, the photographing device 28 includes a camera that photographs an image of the surface of the stored waste 10 in the pit 4, and an illuminator that irradiates light onto the surface of the stored waste 10. This camera is a normal RGB camera. As described below, in this embodiment, the waste quality detection system 2 detects the amount of moisture contained in the waste as a characteristic quantity of the waste. For this reason, in this embodiment, the illuminator can separately irradiate light of a wavelength that is easily absorbed by water (absorption band light) and light of a wavelength that is not easily absorbed by water (non-absorption band light). The camera captures the reflected light of the light irradiated by the illuminator and photographs the surface of the waste. The camera can also photograph the surface of the stored waste 10 using natural light without using the illuminator.

Although not shown, the photographing device 28 may further include a filter changer, and the illuminator may be configured to irradiate light including absorbed band light and non-absorbed band light. The filter changer controls whether the camera is to be fitted with a first bandpass filter that passes only absorbed band light, or a second bandpass filter that passes only non-absorbed band light. It is possible to photograph the stored waste 10 with a camera equipped with the first bandpass filter and to photograph the stored waste 10 with a camera equipped with the second bandpass filter separately.

Although not shown, the photographing device 28 may be configured to include a first camera and a second camera, and the illuminator may be configured to irradiate light including absorbed band light and non-absorbed band light. The first camera is equipped with a first bandpass filter that passes only absorbed band light, and the second camera is equipped with a second bandpass filter that passes only non-absorbed band light. It is possible to photograph the stored waste 10 with the first camera equipped with the first bandpass filter, and to photograph the stored waste 10 with the second camera equipped with the second bandpass filter, either simultaneously or separately.

The configuration of the image capture device 28 is not limited to the above. For example, a spectral camera or an infrared camera may be used as the camera. The image capture device 28 may have various configurations depending on the features to be detected.

The level gauge 30 measures the height of the stored garbage 10 at a predetermined horizontal position in the pit 4. In other words, the level gauge 30 measures the three-dimensional coordinates (x, y, z) of the stored garbage 10 at a predetermined position on the surface of the stored garbage 10. The level gauge 30 is attached to the upper end of the side wall 12. In this embodiment, as shown in FIG. 2, the level gauge 30 is attached to the upper end of the third side wall 12c. The level gauge 30 may be attached to another position. In this embodiment, there is one level gauge 30. A plurality of level gauges 30 may be attached to the pit 4. For example, two level gauges 30 may be attached to the pit 4, one level gauge 30 measures the height of the stored garbage 10 located in half of the virtual area 14 of the entire virtual area 14, and the other level gauge 30 may measure the height of the stored garbage 10 located in the remaining half of the virtual area 14.

In this embodiment, the level meter 30 is a laser distance meter. The level meter 30 irradiates a laser beam toward the garbage surface, measures the time it takes for the laser beam to reflect off the garbage surface and return, and measures the distance d to that position. If the angle that the irradiated laser beam makes with the x-axis of the reference coordinate system is θx, the angle that it makes with the y-axis is θy, and the angle that it makes with the z-axis is θz, the distance on each coordinate axis from the position of the level meter 30 to this surface position is (d x cos θx, d x cos θy, d x cos θz). From this distance and the coordinates of the level meter 30, the three-dimensional coordinates of a specified position on the garbage surface can be obtained.

The level meter 30 uses a pulsed laser beam to radially scan a single plane, as shown by the arrow S in Figure 2. Furthermore, the level meter 30 can rotate so that the laser beam can be emitted in the vertical direction. This allows the level meter 30 to measure the three-dimensional coordinates of the entire surface of the stored waste 10 in the pit 4.

The first calibration device 32 performs calibration to associate a position in the pit 4 with coordinates in the image data captured by the imaging device 28. In this embodiment, coordinates in the image data are associated with coordinates on the reference coordinate axis. In detail, the first calibration device 32 determines the parameters of a conversion equation that converts coordinates on the reference coordinate axis into coordinates in the image data captured by the imaging device 28. The conversion equation with the determined parameters is used in the coordinate conversion unit 42 of the boundary calculation device 36, which will be described later.

この式で、座標（ｘｂ、ｙｂ、ｚｂ）に乗じられている行列は、基準座標軸と撮影装置２８の座標軸との間の回転を補正する回転行列であり、付加されているベクトルは、これらの間の原点ずれを補正する並進ベクトルである。さらに、撮影装置２８での座標（ｘｃ、ｙｃ、ｚｃ）から画像データの座標（ｕ、ｖ）への変換は、以下の式（２）で表される。

第一キャリブレーション装置３２は、式（１）の回転行列のパラメータ（ｒ１１、ｒ１２、・・、ｒ３３）、並進ベクトルのパラメータ（ｔ１、ｔ２、ｔ３）及び式（２）の変換行列のパラメータ（ｆｘ、ｆｙ、Ｃｘ、Ｃｙ）を決定する。 The conversion from the coordinates (xb, yb, zb) on the reference coordinate axes to the coordinates (xc, yc, zc) on the image capturing device 28 is expressed by the following equation (1).

In this formula, the matrix multiplied by the coordinates (xb, yb, zb) is a rotation matrix that corrects the rotation between the reference coordinate axes and the coordinate axes of the image capture device 28, and the added vector is a translation vector that corrects the origin shift between them. Furthermore, the conversion from the coordinates (xc, yc, zc) in the image capture device 28 to the coordinates (u, v) of the image data is expressed by the following formula (2).

The first calibration device 32 determines the parameters (r11, r12, ..., r33) of the rotation matrix in equation (1), the parameters (t1, t2, t3) of the translation vector, and the parameters (fx, fy, Cx, Cy) of the transformation matrix in equation (2).

When determining parameters in the first calibration device 32, a coordinate set (Pi, Qi) (i = 1, 2, ...) is prepared for multiple positions (measurement points) in the pit 4, with the coordinates Pi (xbi, ybi, zbi) in the reference coordinate system and the coordinates Qi (ui, vi) in the image data when the measurement points are photographed by the photographing device 28.

Figure 4 shows how a set of coordinates (Pi, Qi) is prepared. In this embodiment, the connection point M between the bucket 24 of the crane 6 and the rope 22 is set as the measurement point. The drive controller 26 moves the connection point M to a position whose coordinate is P1 on the reference coordinate axis, and the camera 28 photographs the pit 4 at this time. Image data at this location is obtained. The connection point M is then moved to a position whose coordinate is P2 on the reference coordinate axis, and a similar measurement is performed. By repeating this process, multiple image data are obtained.

The position of the connection point M in each image data is measured, and the coordinate Qi in the image data of the measurement point whose coordinate on the reference coordinate axis is Pi is obtained. A set of coordinates (Pi, Qi) (i = 1, 2, ...) is prepared for multiple measurement points in the pit 4. In this embodiment, a set of coordinates is prepared for 10 measurement points.

In this embodiment, the coordinates of connection point M in the image data are measured by software. For example, connection point M is colored a different color from its surroundings, and the position of this color is searched for by software, thereby obtaining the coordinates of connection point M. The coordinates of connection point M in the image data may also be measured by an operator from the image data.

In the above embodiment, the drive controller 26 of the crane 6 moves the connection point M to the position of the coordinate Pi, thereby obtaining the coordinate Pi of the measurement point on the reference coordinate axis. The coordinate Pi of the measurement point on the reference coordinate axis may also be obtained by measuring it with the level meter 30. In this case, the second calibration described below must be performed before the first calibration.

In the above embodiment, the connection point M between the bucket 24 and the rope 22 is set as the measurement point. The measurement point is not limited to this position. The measurement point may be, for example, the tip of the bucket 24. The measurement point may be set so that it can be positioned at a desired location within the pit 4.

The first calibration device 32 determines the parameters of the equations (1) and (2) so that the multiple coordinate sets (Pi, Qi) satisfy the relationships of the above equations (1) and (2). A conventional method is used for this determination. In this embodiment, this determination is performed using functions of the OpenCV (Open Source Computer Vision Library). The first calibration device 32 determines the parameters of the conversion equations (1) and (2) for converting coordinates on the reference coordinate axis to coordinates in the image data.

The second calibration device 34 performs calibration to associate positions in the pit 4 with the measurement results of the level meter 30. As described above, the level meter 30 calculates the three-dimensional coordinates of the object on the reference coordinate axes from the distance to the object measured by irradiating the laser beam and the angle the laser beam makes with the reference coordinate axes at this time. The second calibration device 34 associates the measurement results of the level meter 30 with coordinates on the reference coordinates by associating the direction in which the level meter 30 irradiates the laser beam with the angle it makes with each axis of the reference coordinate axes. For this purpose, coordinates measured by the level meter 30 are prepared for multiple positions (measurement points) in the pit 4.

Figure 5 shows how the coordinates of the measurement points are prepared by the level meter 30. In this embodiment, the measurement points are provided on the side wall 12 of the pit 4. In this embodiment, the level meter 30 measures the coordinates of three measurement points A, B, and C on the first side wall 12a and three measurement points D, E, and F on the second side wall 12b. Measurement points A, B, and C may be located anywhere on the first side wall 12a as long as they are not aligned on a straight line. Measurement points D, E, and F may be located anywhere on the second side wall 12b as long as they are not aligned on a straight line.

The second calibration device 34 identifies the normal direction (direction parallel to the x-axis) of the first side wall 12a from the coordinates of the measurement points A, B, and C, and identifies the normal direction (direction parallel to the y-axis) of the second side wall 12b from the coordinates of the measurement points D, E, and F. The normal direction (direction parallel to the z-axis) of the floor of the pit 4 is identified as a direction perpendicular to both the direction parallel to the x-axis and the direction parallel to the y-axis. The second calibration device 34 determines the angle between the reference coordinate axis and the coordinate axis of the level meter 30 from the directions of these coordinate axes and the direction of the coordinate axes of the level meter 30. In other words, the second calibration device 34 associates the direction in which the level meter 30 emits the laser beam with the angle it makes with each axis of the reference coordinate axis. Using this, the level meter 30 can output the coordinates of the garbage used by the boundary calculation device 36 in the coordinates of the reference coordinate system.

The method of providing the measurement points is not limited to the above. Although not shown, for example, two measurement points G and H having the same height may be provided on the first side wall 12a, and two measurement points I and J having the same height may be provided on the second side wall 12b. Marks of these measurement points are attached to the side wall 12 in advance, and the level meter 30 measures the coordinates of these marks. The second calibration device 34 identifies a direction parallel to the y axis from the vector of the difference between the coordinates of measurement point G and the coordinates of measurement point H, and identifies a direction parallel to the x axis from the vector of the difference between the coordinates of measurement point I and the coordinates of measurement point J. The direction parallel to the z axis can be identified as a direction perpendicular to both the direction parallel to the x axis and the direction parallel to the y axis. By identifying the directions of these coordinate axes, the second calibration device 34 associates the direction in which the level meter 30 emits a laser beam with the angle it makes with each axis of the reference coordinate axis.

The boundary calculation device 36 calculates the boundary of the virtual area 14 on the surface of the stored waste 10 from the image data captured by the photographing device 28 and the height of the waste at a specified position measured by the level meter 30. As shown in FIG. 3, the boundary calculation device 36 includes a boundary point setting unit 40, a coordinate conversion unit 42, and a boundary setting unit 44.

For each virtual area 14, the boundary point setting unit 40 sets a plurality of points (boundary points) to represent the virtual area 14 on the surface of the stored waste 10. In this embodiment, the vertices of the virtual area 14 on the surface of the stored waste 10 are set as the boundary points. For example, when the coordinates of the four vertices of the virtual area 14 on the xy plane are (x1, y1), (x2, y1), (x2, y2), and (x1, y2), and the heights of the respective vertices on the surface of the stored waste 10 measured by the level meter 30 are z1, z2, z3, and z4, the four points with coordinates (x1, y1, z1), (x2, y1, z2), (x2, y2, z3), and (x1, y2, z4) are set as the boundary points.

Figure 6 shows how boundary points are set. Figure 6(a) is an example of image data of stored waste 10 photographed by the photographing device 28, and Figure 6(b) is a view of the stored waste 10 viewed on the x-z plane at the position of the dividing line of the virtual area 14 parallel to the x-axis. In Figure 6(b), the two-dot chain line represents the position of the dividing line parallel to the y-axis. In Figure 6(b), symbols p1 and p2 represent the boundary points of the virtual area 14 located at the leftmost position in the figure. Symbols p2 and p3 represent the boundary points of the virtual area 14 located second from the left in the figure.

In the coordinate conversion unit 42, the coordinates of each boundary point on the reference coordinate axis are converted into coordinates in the image data. In this conversion, equations (1) and (2) whose parameters have been set by the first calibration device 32 are used. As a result, the coordinates (u1, v1), (u2, v2), (u3, v3), and (v4, v4) of the boundary points in the image data are obtained. This conversion is performed in each virtual area 14. Figure 6(a) shows the boundary points in the image data obtained for some of the virtual areas 14. In Figure 6(a), the vertices of the rectangle drawn superimposed on the surface of the stored waste 10 are the boundary points in the image data obtained by the above conversion.

The boundary setting unit 44 sets the boundary of the virtual area 14 in the image data by connecting boundary points in the image data. In FIG. 6(a), the boundary of the virtual area 14 set in the image data for some of the virtual areas 14 is shown.

In the above embodiment, the four vertices of the virtual area 14 on the surface of the stored waste 10 are set as boundary points. The number of boundary points may be increased. For example, the midpoints of adjacent vertices among these four vertices may be added as boundary points. In this case, the number of boundary points is eight, and their coordinates are (x1, y1, z1), ((x1 + x2) / 2, y1, z5), (x2, y1, z2), (x2, (y1 + y2) / 2, z6), (x2, y2, z3), ((x1 + x2) / 2, y2, z7), (x1, y2, z4), (x1, (y1 + y2) / 2, z8). Here, z5, z6, z7, and z8 are the heights of the stored waste 10 at the added midpoints, respectively. The coordinate conversion unit 42 converts these coordinates, and the boundary setting unit 44 connects the boundary points in the image data to determine the virtual area 14 in the image data.

The position of the added boundary point may be changed depending on the shape of the surface of the stored waste 10 in the virtual area 14. For example, the highest point or the lowest point between adjacent vertices may be added as a boundary point.

The number of boundary points may be changed depending on the shape of the surface of the stored waste 10 in the virtual area 14. For example, in a virtual area 14 where the height difference of the unevenness on the surface of the stored waste 10 is greater than a predetermined value, the number of boundary points may be the above-mentioned 8, and in a virtual area 14 where the height difference of the unevenness is equal to or less than this predetermined value, the number of boundary points may be the above-mentioned 4.

The feature amount calculation device 38 calculates the feature amount of the garbage for each virtual area 14 in the image data. In other words, the virtual area 14 is the evaluation unit when calculating the feature amount of the garbage. In this embodiment, the moisture content of the garbage is detected as the feature amount of the garbage. In this embodiment, the feature amount calculation device 38 uses image data (moisture evaluation image) taken by irradiating the stored garbage 10 in the pit 4 with absorption band light and image data (reference image) taken by irradiating non-absorption band light. In the moisture evaluation image, garbage with a high moisture content absorbs a lot of light, so the image is dark. In the reference image, the amount of light absorbed by moisture is small, so the image is bright even for garbage with a high moisture content. The boundary calculation device 36 determines the virtual area 14 for each image data of the moisture evaluation image and the reference image.

As the moisture evaluation image, image data captured with a camera equipped with a first bandpass filter may be used, and as the reference image, image data captured with a camera equipped with a second bandpass filter may be used. In these captures, the illuminator irradiates light that includes both absorbed and non-absorbed band light.

In this embodiment, the feature calculation device 38 calculates the average brightness Lw in the moisture evaluation image and the average brightness Lr in the reference image for each virtual region 14. The difference between these (Lr-Lw) is taken as the feature representing the moisture content of this virtual region 14. The greater this difference (Lr-Lw), the greater the moisture content is determined to be. The greater this difference, the lower the quality of the garbage is considered to be.

The method of calculating the feature representing the moisture content is not limited to the above. The ratio (Lw/Lr) of the average brightness Lw to the average brightness Lr may be used as the feature representing the moisture content of this virtual region 14. The smaller this ratio (Lw/Lr), the greater the moisture content is determined to be. In addition, the attenuation rate ((Lr-Lw)/Lr) of the average brightness of the moisture evaluation image relative to the reference image may be used as the feature representing the moisture content of this virtual region 14. The greater this attenuation rate ((Lr-Lw)/Lr), the greater the moisture content is determined to be.

A method for detecting litter quality using the litter quality detection system 2 will be described below.
(A) a calibration step of the imaging device 28;
(B) a calibration process for the level meter 30; and (C) a garbage quality detection process.
The step (A) is normally performed when the imaging device 28 is installed, when the state of the imaging device 28 changes, such as when a part is replaced. The step (B) is normally performed when the level meter 30 is installed, when the state of the level meter 30 changes, such as when a part is replaced. Steps (A) and (B) are preparatory steps for detecting the garbage quality with high accuracy. Step (C) is performed every time the garbage quality is detected.

In the above step (A), a first calibration device 32 is used. This step includes the following steps:
The method includes (A1) using an imaging device 28 to obtain image data for each of a plurality of different positions of a specific measurement point of the crane 6 in the pit 4; and (A2) performing a calibration process to match the coordinates in the image data with the coordinates on a reference coordinate axis based on the coordinates of the measurement point in the image data.

In the above step (A1), the measurement point is positioned at a location whose coordinate is P1 on the basic coordinate axes and photographed by the photographing device 28 to obtain image data. The measurement point is moved to a location whose coordinate is P2 on the basic coordinate axes and photographed by the photographing device 28 to obtain different image data. By repeating this process, image data can be obtained when the measurement point on the crane 6 is positioned at multiple different locations within the pit 4.

In the above step (A2), the coordinates of the measurement points are measured from each image data to obtain coordinates Qi. This results in multiple coordinate sets (Pi, Qi). The parameters of equations (1) and (2) are determined using OpenCV functions so that the coordinate set (Pi, Qi) satisfies the above equations (1) and (2).

In the above step (A1), one measurement point may be photographed, and the coordinate Qi may be obtained from the image data, after which the next measurement point may be photographed. In this case, in the above step (A2), a process is carried out to determine the parameters of equations (1) and (2).

The coordinates of the measurement point in the reference coordinate system may be measured by the level meter 30. In this case, in step (A1), the pit 4 is photographed by the photographing device 28, and the coordinates of the measurement point are measured by the level meter 30. In the above step (A2), a process is carried out to determine the parameters of equations (1) and (2).

In the above step (B), the second calibration device 34 is used. This step is as follows:
(B1) a step of measuring the coordinates of a plurality of measurement points located on the first side wall 12a and a plurality of measurement points located on the second side wall 12b perpendicular to the first side wall 12a using a level meter 30; and (B2) a step of calibrating the level meter 30 based on the above measurement results to correspond the measurement results to coordinates on a reference coordinate axis.

In the above step (B2), the direction in which the level meter 30 emits the laser beam and the angle this makes with each axis of the reference coordinate system are associated with the coordinates of the measurement point measured by the level meter 30 in step (B1).

In the above step (C), the quality of the garbage 10 stored in the pit 4 is detected. This step is as follows:
The process includes: (C1) photographing the surface of the stored waste 10 to obtain image data; (C2) measuring the height of the stored waste 10; (C3) setting boundary points to represent virtual areas 14 on the surface of the stored waste 10; (C4) converting the coordinates of the boundary points into coordinates in the image data; (C5) setting the boundaries of the virtual areas 14 in the image data; and (C6) detecting the characteristics of the waste for each virtual area 14 in the image data.

Either of the above steps (C1) and (C2) can be carried out first. However, it is preferable to carry out steps (C1) and (C2) with as little time between them as possible. It is preferable to carry out steps (C1) and (C2) with a time interval such that the state of the garbage in pit 4 does not change between them.

In the above step (C1), the surface of the stored waste 10 is photographed using the photographing device 28. In this embodiment, a moisture evaluation image photographed by irradiating the waste with absorption band light and a reference image photographed using non-absorption band light are obtained.

In the above step (C2), the height of the stored waste 10 at a specified position is measured using a level meter 30. The coordinates of the surface of the stored waste 10 are obtained.

In the above step (C3), the boundary calculation device 36 is used. The vertices of the virtual area 14 on the surface of the stored waste 10 are set as boundary points.

In step (C4) above, the coordinates of each boundary point on the reference coordinate axis are converted to coordinates in the image data using the above-mentioned formulas (1) and (2).

In the above step (C5), the boundary of the virtual region 14 in the image data is set by connecting the boundary points in the image data for each of the moisture assessment image and the reference image.

In the above step (C6), the average brightness in the moisture assessment image and the average brightness in the reference image are calculated within the boundary of each virtual region 14. The average brightness in the moisture assessment image and the average brightness in the reference image are compared. As a result, the moisture content of each virtual region 14 is detected as a feature.

The garbage quality detected by the garbage quality detection system 2 is sent to the drive controller 26 of the crane 6 and is used by the crane 6 to agitate the stored garbage 10. For example, the crane 6 grabs garbage in the virtual area 14 with a high moisture content and drops it into the virtual area 14 with a low moisture content. At this time, for example, the garbage in the virtual area 14 with the highest moisture content is dropped into the virtual area 14 with the lowest moisture content, and the garbage in the virtual area 14 with the second highest moisture content is dropped into the virtual area 14 with the second lowest moisture content. The garbage quality detection system 2 performs the above steps (C1)-(C6) on the agitated garbage, and the garbage quality is detected. Detection of garbage quality by the garbage quality detection system 2 and agitation by the crane 6 are repeated a predetermined number of times. By repeating the detection of garbage quality and agitation, the garbage quality is homogenized.

The effects of the present invention are explained below.

In the litter quality detection system 2 according to the present invention, the boundaries of the virtual areas 14 in the image data are calculated from the image data of the stored litter 10 photographed by the photographing device 28 and the height of the stored litter 10 at a specified horizontal coordinate measured by the level meter 30. The characteristics of the litter are calculated from the image data within each virtual area 14. In this detection system 2, the boundaries of the virtual areas 14 in the image data are calculated and the characteristics of the litter are calculated within these virtual areas 14, so that the areas in the image data where the characteristics of the litter are calculated correspond with high precision to the areas on the surface of the actual stored litter 10. By using this litter quality detection system 2, efficient uniformity of litter quality can be achieved.

In this embodiment, the vertices of the virtual area 14 on the surface of the stored waste 10 are taken as boundary points, and the coordinates of the boundary points in the image data of the virtual area 14 are calculated based on the coordinates of the boundary points. The boundary of the virtual area 14 in the image data is calculated from the coordinates of the boundary points in the image data. This method is simple. With this method, the boundary of the virtual area 14 in the image data can be calculated efficiently.

In this embodiment, a predetermined position of the crane 6 is set as the measurement point, and this measurement point is positioned at several different locations within the pit 4 and photographed by the photographing device 28. A set is prepared of the coordinates of the measurement point in the image data obtained in this way and the coordinates of this measurement point on the reference coordinate axis. Based on these sets of coordinates, parameters are determined for matching the coordinates on the reference coordinate axis with the coordinates in the image data. In this garbage quality detection system 2, the measurement point can be set at a desired position within the pit 4. With this method, the coordinates on the reference coordinate axis can be matched with the coordinates in the image data with high accuracy.

In this embodiment, a specific position of the crane 6 is set as the measurement point, so the coordinate of the measurement point on the reference coordinate axis can be grasped by the drive controller 26 of the crane 6. This makes it easy to obtain the coordinate of the measurement point on the reference coordinate axis.

The coordinates of the measurement point on the reference coordinate axis may be measured by the level meter 30. This allows the coordinates of the measurement point on the reference coordinate axis to be obtained with high accuracy.

In this embodiment, the coordinates of a specific measurement point provided on the first side wall 12a of the pit 4 and a specific measurement point provided on the second side wall 12b perpendicular to the first side wall 12a are measured by the level meter 30, and based on the coordinates of these points on the reference coordinate axis, the coordinates on the level meter 30 are associated with the coordinates on the reference coordinate axis. This method makes it possible to easily and accurately associate the coordinates on the level meter 30 with the coordinates on the reference coordinate axis.

[Second embodiment]
Fig. 7 is a block diagram showing a litter type detection system 50 according to another embodiment of the present invention. This litter type detection system 50 is the same as the litter type detection system 2 of the embodiment in Fig. 3, except for a boundary calculation device 52. The boundary calculation device 52 will be described below. As shown in Fig. 7, the boundary calculation device 52 includes a virtual area average height calculation unit 54, a boundary point setting unit 56, a coordinate conversion unit 58, and a boundary setting unit 60.

The virtual area average height calculation unit 54 calculates the average height of the stored waste 10 in each virtual area 14 from the coordinates of the surface of the stored waste 10 measured by the level meter 30.

In the boundary point setting unit 56, for each virtual area 14, the vertex of the virtual area 14 at the average height position of the virtual area 14 is set as the boundary point. Figure 8 shows how the boundary points are set. Figure 8(a) is an example of image data of stored waste 10 photographed by the photographing device 28, and Figure 8(b) is a view of the stored waste 10 as seen in the x-axis-z-axis plane at the position of the dividing line of the virtual area 14 that is parallel to the x-axis. The symbol H in Figure 8(b) represents a horizontal plane having the average height of the virtual area 14. The vertex of the virtual area 14 at the position of this plane H is set as the boundary point. For example, if the coordinates of the vertices of one virtual area 14 on the surface of the stored waste 10 are (x1, y1, z1), (x2, y1, z2), (x2, y2, z3), and (x1, y2, z4), and the average height of the stored waste 10 in this virtual area 14 is za, then four points with coordinates (x1, y1, za), (x2, y1, za), (x2, y2, za), and (x1, y2, za) are set as boundary points. In FIG. 8(b), symbols p1 and p2 represent the boundary points of the virtual area 14 located at the leftmost position in the figure. Symbols p3 and p4 represent the boundary points of the virtual area 14 located second from the left in the figure.

In the coordinate conversion unit 58, the coordinates of each boundary point are converted into coordinates in the image data. This conversion is performed using the above formulas (1) and (2). As a result, the coordinates (u1, v1), (u2, v2), (u3, v3), and (v4, v4) in the image data of the boundary points are obtained. Figure 8(a) shows the boundary points in the obtained image data for a portion of the virtual area 14. In Figure 8(a), the vertices of the rectangle drawn superimposed on the surface of the stored waste 10 are the boundary points in the image data obtained by the above conversion.

The boundary setting unit 60 sets the boundary of the virtual area 14 in the image data by connecting the boundary points in the image data. Figure 8 (a) shows the boundary of the virtual area 14 set in the image data for some of the virtual areas 14.

In the above embodiment, the average height of the garbage in the virtual region 14 is calculated, and for each virtual region 14, the vertex of the virtual region 14 at the average height position of this virtual region 14 is set as the boundary point. The median height of the garbage in the virtual region 14 (the median height between the highest and lowest points) may be calculated, and for each virtual region 14, the vertex of the virtual region 14 at the median height position of this virtual region 14 may be set as the boundary point.

In this embodiment, for each virtual area 14, the average height or median height of the stored waste 10 in this virtual area 14 on the reference coordinate axis is calculated, and the coordinates of the boundary points in the image data are calculated based on the vertices of the virtual area 14 at this height. The boundary of the virtual area 14 in the image data is calculated from the coordinates of the boundary points in the image data. This method is simple. With this method, the boundary of the virtual area 14 in the image data can be calculated efficiently.

As explained above, with this litter quality detection system, the area in the image data where the litter features have been calculated can be easily and accurately matched to the actual litter surface area. By using this litter quality detection system, it is possible to efficiently standardize litter quality. This clearly shows the superiority of the present invention.

The waste quality detection system described above can be applied to a variety of waste storage facilities.

Reference Signs List 1: Waste storage facility 2: Waste quality detection system 4: Pit 6: Crane 8: Hopper 10: Stored waste 12: Side wall 14: Virtual area 22: Rope 24: Bucket 26: Drive controller 28: Photography device 30: Level gauge 32: First calibration device 34: Second calibration device 36, 52: Boundary calculation device 38: Feature calculation device 40, 56: Boundary point setting unit 42, 58: Coordinate conversion unit 44, 60: Boundary setting unit

Claims (8)

A garbage quality detection system for garbage stored in a pit, comprising:
an imaging device for imaging the dust and obtaining image data;
A level meter for measuring the height of the waste at a predetermined horizontal position;
a boundary calculation device that calculates a boundary in the image data of a predetermined virtual area that serves as a unit area when evaluating the quality of litter, based on the horizontal position and height and the image data;
a feature amount calculation device for calculating a feature amount of the dust in the virtual area from the image data within the boundary ,
A litter quality detection system , wherein the boundary calculation device calculates the boundary of the virtual area in the image data based on the horizontal positions and heights of the vertices of the virtual area on the surface of the litter . A garbage quality detection system for garbage stored in a pit, comprising:
an imaging device for imaging the dust and obtaining image data;
A level meter for measuring the height of the waste at a predetermined horizontal position;
a boundary calculation device that calculates a boundary in the image data of a predetermined virtual area that serves as a unit area when evaluating the quality of litter, based on the horizontal position and height and the image data;
a feature amount calculation device for calculating a feature amount of the dust in the virtual area from the image data within the boundary;
Equipped with
A garbage quality detection system, wherein the boundary calculation device calculates the boundary of the virtual area in the image data based on the horizontal position and height of the vertex of the virtual area at the average height position of the garbage in the virtual area. A garbage quality detection system for garbage stored in a pit, comprising:
an imaging device for imaging the dust and obtaining image data;
A level meter for measuring the height of the waste at a predetermined horizontal position;
a boundary calculation device that calculates a boundary in the image data of a predetermined virtual area that serves as a unit area when evaluating the quality of litter, based on the horizontal position and height and the image data;
a feature amount calculation device for calculating a feature amount of the dust in the virtual area from the image data within the boundary;
Equipped with
Further comprising a first calibration device;
A garbage quality detection system in which the first calibration device performs calibration to correspond the coordinates in the image data to the position in the pit based on image data captured by the photographing device by positioning a specific measurement point of a crane used in the pit at multiple different locations within the pit. A garbage quality detection system for garbage stored in a pit, comprising:
an imaging device for imaging the dust and obtaining image data;
A level meter for measuring the height of the waste at a predetermined horizontal position;
a boundary calculation device that calculates a boundary in the image data of a predetermined virtual area that serves as a unit area when evaluating the quality of litter, based on the horizontal position and height and the image data;
a feature amount calculation device for calculating a feature amount of the dust in the virtual area from the image data within the boundary;
Equipped with
Further comprising a second calibration device;
A garbage quality detection system in which the second calibration device performs calibration to correspond the measurement results of the level meter to a position in the pit based on the horizontal coordinate and height of a specified measurement point on a first side wall of the pit and the horizontal coordinate and height of a specified measurement point on a second side wall perpendicular to the first side wall , measured by the level meter. 5. The waste quality detection system according to claim 1, wherein the characteristic amount is a moisture content of the waste. A method for detecting the quality of garbage stored in a pit, comprising the steps of:
A step of photographing the surface of the rubbish with an imaging device to obtain image data;
measuring the height of the waste at a predetermined horizontal position with a level meter;
The method includes a step of calculating a boundary in the image data of a predetermined virtual area that is a unit area for evaluating the quality of litter based on the horizontal position and height and the image data, and a step of obtaining a feature amount of litter in this virtual area from the image data within the boundary ,
The method further includes a step of calibrating the image capture device,
The calibration step of the imaging device includes:
A process of positioning a predetermined measurement point of the crane at a plurality of different locations in the pit and photographing the measurement point using the photographing device to obtain image data.
and
A step of performing a calibration for associating the coordinates in the image data with the positions in the pits based on the coordinates in the image data of the measurement points.
A method for detecting litter quality, comprising : A method for detecting the quality of garbage stored in a pit, comprising the steps of:
A step of photographing the surface of the rubbish with an imaging device to obtain image data;
measuring the height of the waste at a predetermined horizontal position with a level meter;
A step of calculating a boundary in the image data of a predetermined virtual area that is a unit area for evaluating the quality of the waste, based on the horizontal position and height and the image data.
and
A step of determining a feature amount of dust in the virtual area from image data within the boundary.
Including,
The method further includes a step of calibrating the level meter,
The step of calibrating the level meter comprises:
A garbage quality detection method comprising: a step of measuring, using the level meter, the horizontal coordinate and height of a specified measurement point on a first side wall of the pit, and the horizontal coordinate and height of a specified measurement point on a second side wall perpendicular to the first side wall; and a step of performing calibration based on the results of the measurement to correspond the measurement results of the level meter to a position in the pit. The facility is equipped with a pit for storing garbage, a crane for stirring the garbage, and a garbage quality detection system.
The waste quality detection system comprises:
an imaging device for imaging the dust and obtaining image data;
A level meter for measuring the height of the waste at a predetermined horizontal position;
a boundary calculation device that calculates a boundary in the image data of a predetermined virtual area that serves as a unit area when evaluating the quality of litter, based on the horizontal position and height and the image data;
a feature amount calculation device for calculating a feature amount of dust in the virtual area from image data within the boundary,
The crane agitates the waste based on the characteristic amount of the waste ,
A waste storage facility , wherein the boundary calculation device calculates the boundary of the virtual area in the image data based on horizontal positions and heights of vertices of the virtual area on the surface of the waste .

Images