JP7057915B2 - Object detection system, object detection method - Google Patents

Description

The present invention relates to an object detection system and an object detection method for detecting a predetermined object in an image.

For example, in a physical store such as a convenience store or a supermarket, products may be displayed based on a shelf allocation contracted in advance between a product manufacturer or wholesaler and a retailer.

Here, in order for the product manufacturer or wholesaler to confirm whether or not the product is displayed according to the contracted shelving allocation, the product manufacturer or wholesaler needs to visit the actual store. , There was a risk that the time and effort required to check the display status of the products would increase.

Therefore, in order to improve the efficiency of the work of checking the display status, a technique for detecting the image area of the product included in the image data as the detection image area from the image data obtained by photographing the product shelf on which the product (object) is displayed. Is disclosed (for example, Patent Document 1).

In the technique described in Patent Document 1, for example, an image of a product such as "can" or "PET bottle" is used by a learning method using teacher data prepared for each category such as "can top" or "PET bottle". The area is detected as a detection image area.

JP-A-2015-210651

By the way, when a product is detected from image data by a learning method, it is necessary to prepare a large amount of teacher data in order to improve the detection accuracy, which may increase the cost of collecting teacher data. Further, in order to perform the learning process, a long time of several hours to several days and a high-load calculation process are required, which may increase the cost of the product detection process.

The present invention has been made in view of the above problems, and an object of the present invention is to provide an object detection system and an object detection method capable of detecting a predetermined object in an image by a simple process.

In order to solve the above problems, first of all, in the present invention, the acquisition means for acquiring an image of a predetermined object, the horizontal direction of the acquired image is the X-axis direction, and the vertical direction of the image is the Y-axis direction. When, the number of pixels whose brightness is equal to or less than a predetermined threshold among the images arranged in one of the X-axis and the Y-axis is counted for each coordinate of the other of the X-axis and the Y-axis, and the count. Provided is an object detection system including a detection means for detecting the predetermined object by extracting the other coordinate which is a boundary of the predetermined object in the image based on the number of pixels. (Invention 1).

According to the present invention (Invention 1), among the pixels arranged in one of the X-axis and the Y-axis, the number of pixels whose brightness is equal to or less than a predetermined threshold value (that is, dark pixels whose brightness is equal to or less than the threshold value) is based on the number of pixels. A predetermined object can be detected by extracting coordinates that are boundaries of a predetermined object from the coordinates of the other of the X-axis and the Y-axis. This makes it possible to detect a predetermined object by counting the number of pixels whose brightness is equal to or less than a predetermined threshold value. A predetermined object can be detected by a simple process, and further, the cost required for the detection process can be suppressed.

In the above invention (invention 1), the counting means may count the number of the pixels in a state where the acquired image is rotated by a predetermined rotation angle (invention 2).

According to the present invention (Invention 2), when the shape of a predetermined object in an image does not follow any direction of the X-axis and the Y-axis (for example, the predetermined object is in either the X-axis or the Y-axis direction). Even if it has a shape that does not follow, or if a predetermined object is imaged at an angle to the X-axis and Y-axis, etc.), by rotating the image, the predetermined object in the image Since the coordinates that are the boundaries of a predetermined object can be extracted in a state where the shape of the object is along either the X-axis or the Y-axis direction, the detection accuracy of the predetermined object can be improved.

In the above inventions (Inventions 1 and 2), the counting means may count the number of the pixels by using each of the plurality of threshold values when a plurality of the threshold values are provided (Invention 3).

According to the present invention (Invention 3), the number of pixels counted when a large number of pixels having higher brightness are arranged in one of the X-axis and Y-axis directions, and the pixel having lower brightness is the one. Since it is possible to increase the difference from the number of pixels counted when many pixels are lined up in the direction, it is possible to improve the extraction accuracy of the coordinates that are the boundaries of a predetermined object.

In the above invention (Invention 3), the counting means may weight the number of the pixels when counting using any of the plurality of thresholds according to the value of the threshold. Good (Invention 4).

According to the present invention (Invention 4), for example, when the number of pixels is heavily weighted as the threshold value becomes smaller, a large number of pixels having higher brightness are arranged in one of the X-axis and the Y-axis. The boundary between a given object can be made larger because the difference between the number of pixels counted in the case and the number of pixels counted when more pixels with lower brightness are lined up in one direction is possible. It is possible to further improve the extraction accuracy of the coordinates.

In the above inventions (Inventions 1 to 4), the detection means is one of the other coordinates when a plurality of the other coordinates satisfying a predetermined condition regarding the number of pixels are present within a predetermined range. May be extracted as the boundary of a predetermined object (Invention 5).

According to the invention (Invention 5), for example, when a plurality of other coordinates satisfying a predetermined condition regarding the number of pixels are within a range smaller than the size of a predetermined object, each of the other coordinates is predetermined. Since it is possible to suppress erroneous extraction as the boundary of the object, it is possible to improve the detection accuracy of a predetermined object.

In the above inventions (Inventions 1 to 5), when the predetermined object exists in each of the plurality of areas in the image, each of the plurality of areas is specified by performing image processing on the image. The counting means counts the number of the pixels in at least one of the plurality of areas, and the detecting means detects the predetermined object existing in the at least one area. It may be (Invention 6).

According to the present invention (Invention 6), even when a predetermined object exists in each of a plurality of areas in an image, the predetermined object in each area can be detected.

Secondly, the present invention is X when the computer acquires an image of a predetermined object and the horizontal direction of the acquired image is the X-axis direction and the vertical direction of the image is the Y-axis direction. Based on the step of counting the number of pixels whose brightness is equal to or less than a predetermined value among the pixels arranged in one of the axes and the Y-axis for each coordinate of the other of the X-axis and the Y-axis, and the number of counted pixels. Provided is an object detection method for executing each step of detecting the predetermined object by extracting the other coordinate which is the boundary of the predetermined object in the image (Invention 7). ..

According to the object detection system and the object detection method of the present invention, a predetermined object in an image can be detected by a simple process.

It is a figure which shows schematic the basic structure of the object detection system which concerns on one Embodiment of this invention. It is a block diagram which shows the structure of a detection device. It is a functional block diagram for demonstrating the function which plays a main role in the object detection system which concerns on embodiment. It is a flowchart which shows an example of the process of specifying ROI in an image. (A) to (d) are diagrams for explaining an example of images before and after image processing. It is a flowchart which shows an example of the process of detecting an object in ROI. It is a figure which shows the structural example of the count data. (A) to (b) are diagrams illustrating an example of ROI before and after rotation. It is a figure explaining an example of ROI after object detection. It is a flowchart which shows an example of the main processing of the object detection system which concerns on embodiment.

Hereinafter, the object detection system and the object detection method according to the embodiment of the present invention will be described in detail with reference to the accompanying drawings. However, this embodiment is an example, and the present invention is not limited thereto.

(1) Basic Configuration of Object Detection System FIG. 1 is a diagram schematically showing a basic configuration of an object detection system according to an embodiment of the present invention. As shown in FIG. 1, the object detection system according to the present embodiment includes a detection device 10, and the detection device 10 is a product imaged by using a communication terminal 20 in a real store such as a convenience store or a supermarket. When the image of the shelf S is acquired from the communication terminal 20, a predetermined object (in the present embodiment, the product displayed on the product shelf S) in the image is detected.

In this embodiment, the case where the product displayed on the product shelf S is a predetermined object will be described as an example, but the predetermined object may be any object that can be detected. good.

The detection device 10 and the communication terminal 20 are connected to a communication network NW (network) such as the Internet or a LAN (Local Area Network), and can communicate with each other via the communication network NW.

The communication terminal 20 is, for example, a terminal device having a function of capturing a moving image or a still image, for example, a mobile terminal, a smartphone, a PDA (Personal Digital Assistant), a personal computer, and a television having a two-way communication function. It may be a terminal device operated by an individual user, such as a receiver (including a so-called multifunctional smart TV).

In the present embodiment, a case where the communication terminal 20 captures an image of the product shelf S and transmits the image to the detection device 10 will be described as an example. However, the communication terminal 20 captures an image of, for example, a digital camera or a digital video camera. It may be a device for transferring the image of the product shelf S captured by the device to the detection device 10.

(2) Configuration of Detection Device 10 The detection device 10 will be described with reference to FIG. FIG. 2 is a block diagram showing an internal configuration of the detection device 10. As shown in FIG. 2, the detection device 10 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a non-volatile memory 14, and a display processing unit 15. , A display unit 16, an input unit 17, and a communication interface unit 18, and a bus 19 for transmitting a control signal or a data signal between the units is provided. The detection device 10 may be, for example, a general-purpose personal computer.

When the power is turned on to the detection device 10, the CPU 11 loads various programs stored in the ROM 12 or the non-volatile memory 14 into the RAM 13 and executes them. In the present embodiment, the CPU 11 reads and executes a program stored in the ROM 12 or the non-volatile memory 14, and thereby obtains means 31, specifying means 32, counting means 33, and detecting means 34 (shown in FIG. 3), which will be described later. To realize the function of.

The non-volatile memory 14 is, for example, an HDD (Hard Disk Drive), and stores a program executed by the CPU 11 and data referenced by the CPU 11. Further, count data (shown in FIG. 7), which will be described later, is stored in the non-volatile memory 14.

The display processing unit 15 displays the display data given by the CPU 11 on the display unit 16. The display unit 16 is, for example, an LCD (Liquid Crystal Display) monitor including thin film transistors arranged in a matrix in pixel units, and by driving the thin film transistors based on the display data, the displayed data is displayed on the display screen. indicate.

The input unit 17 is an information input device such as a mouse or a keyboard. The communication interface unit 18 includes an interface circuit for performing communication via the communication network NW.

(3) Outline of Each Function in the Object Detection System The functions realized by the object detection system of the present embodiment will be described with reference to FIG. FIG. 3 is a functional block diagram for explaining a function that plays a major role in the object detection system of the present embodiment. In the functional block diagram of FIG. 3, the acquisition means 31, the counting means 33, and the detecting means 34 correspond to the main configurations of the present invention. Although the specific means 32 is not an essential configuration, it is a component for further favoring the present invention.

The acquisition means 31 has a function of acquiring an image of an image of a product (predetermined object) displayed on the product shelf S. The function of the acquisition means 31 is realized, for example, as follows.

When the CPU 11 of the detection device 10 receives (acquires) the image of the product shelf S transmitted from the communication terminal 20 via the communication interface unit 18, the CPU 11 stores the received image in, for example, the RAM 13 or the non-volatile memory 14.

When the product (predetermined object) displayed on the product shelf S exists in each of the plurality of areas in the image, the specific means 32 performs image processing on the image to each of the plurality of areas. It has a function to identify. Thereby, even if the product exists in each of the plurality of areas in the image, the product in each area can be detected. In this embodiment, a case where each of a plurality of areas separated by a plurality of shelves of the product shelf S is specified in the image will be described as an example.

The function of the specific means 32 is realized as follows, for example, with reference to the flowchart of FIG. 4 and FIG. FIG. 5 is a diagram illustrating an example of an image before and after image processing. The CPU 11 of the detection device 10 first performs preprocessing on the image (step S100). Specifically, the CPU 11 smoothes the image (shown in FIG. 5A) acquired by the function of the acquisition means 31 by grayscale and then performing, for example, Gaussian blur processing.

Next, the CPU 11 binarizes the smoothed image and then extracts the edge using an edge extraction algorithm such as the Canny method (step S102). The image after edge extraction is shown, for example, as shown in FIG. 5 (b).

Next, the CPU 11 detects a straight line extending in the horizontal direction (X-axis direction in the figure) in the image by performing a feature extraction method such as a stochastic Hough transform on the image from which the edge is extracted (the X-axis direction in the figure). Step S104). In this embodiment, it is assumed that each of the plurality of shelves of the product shelf S is detected as a straight line. Here, when detecting a straight line, the CPU 11 has a predetermined value (for example, 10 pixels) in the coordinate difference in the vertical direction (Y-axis direction in the figure) of the line segments extending in the horizontal direction at both ends. Line segments less than may be detected as straight lines. The image after the straight line detection is shown, for example, as shown in FIG. 5 (c). FIG. 5C shows a case where five horizontal straight lines L are detected as an example.

Then, the CPU 11 sets the ROI by dividing the image after the straight line detection into a plurality of areas (ROI; Region of Interest) using the detected straight line L (step S106). Here, the ROI in the present embodiment means an area including a plurality of products (objects). When setting the ROI using a plurality of straight lines L, the CPU 11 determines the vertical direction of the pixels at the ends of the two straight lines L adjacent to each other in the vertical direction of the image (Y-axis direction in the figure) (FIG. When the coordinate difference in the middle Y-axis direction is equal to or greater than a predetermined value, the area sandwiched between the two straight lines L may be set as the ROI. The image after setting the ROI is shown, for example, as shown in FIG. 5 (d). FIG. 5D shows a case where three ROIs are set as an example.

The counting means 33 is among the pixels arranged in the Y-axis direction (one of the X-axis and the Y-axis) when the horizontal direction of the acquired image is the X-axis direction and the vertical direction of the image is the Y-axis direction. It has a function of counting the number of pixels whose brightness is equal to or less than a predetermined threshold for each coordinate of the X-axis (the other of the X-axis and the Y-axis).

Further, the counting means 33 counts the number of pixels for at least one area among the areas specified by the function of the specific means 32 when the product (predetermined object) exists in each of the plurality of areas in the image. You may.

Further, the counting means 33 may count the number of pixels in a state where the acquired image is rotated by a predetermined rotation angle. In this case, when the shape of the product (predetermined object) in the image does not follow either the X-axis or the Y-axis direction (for example, the product has a shape that does not follow either the X-axis or the Y-axis direction). Even if the product is imaged at an angle to the X-axis and Y-axis, etc.), by rotating the image, the shape of the product in the image can be either X-axis or Y-axis. Since the coordinates that are the boundaries of the products can be extracted along the direction, it is possible to improve the detection accuracy of the products.

Furthermore, when a plurality of threshold values are provided, the counting means 33 may count the number of pixels using each of the plurality of threshold values. In this case, the number of pixels counted when a large number of pixels with higher brightness are arranged in one of the X-axis and Y-axis, and the number of pixels with lower brightness are arranged in a large number in one direction. Since the difference from the number of pixels counted in can be increased, the accuracy of extracting the coordinates that are the boundaries of the product can be improved.

Further, the counting means 33 may weight the number of pixels when counting using any of the plurality of thresholds according to the value of the threshold. In this case, for example, when the number of pixels is heavily weighted as the threshold value becomes smaller, the pixels counted when a large number of pixels having higher brightness are lined up in one of the X-axis and Y-axis directions. Since the difference between the number and the number of pixels counted when many pixels with lower brightness are lined up in one direction can be made larger, the accuracy of extracting the coordinates that are the boundaries of the product is further improved. Can be made to.

The detection means 34 has a function of detecting a product by extracting the coordinates of the X-axis (the other coordinate) that is the boundary of the product (predetermined object) in the image based on the number of counted pixels. ..

Further, when a plurality of X-axis coordinates (the other coordinate) satisfying a predetermined condition regarding the number of pixels exist within a predetermined range, the detection means 34 sets the coordinates of any one X-axis as a product (predetermined). It may be extracted as the boundary of the object). As a result, for example, when a plurality of X-axis coordinates satisfying a predetermined condition regarding the number of pixels are within a range smaller than the size of the product, each of these X-axis coordinates is erroneously extracted as the boundary of the product. It is possible to improve the detection accuracy of the product because it can be suppressed.

The functions of the counting means 33 and the detecting means 34 are realized as follows, for example, with reference to the flowchart of FIG. Here, the number of pixels when the image is rotated at each of a plurality of rotation angles (for example, the rotation angle is −20 ° to + 20 °) is set to a plurality of threshold values (for example, the threshold value is 50 to 200). The case of counting using the above will be described as an example.

As a function of the counting means 33, the CPU 11 of the detection device 10 grayscales one ROI, and in a state where the ROI is rotated by a predetermined rotation angle (for example, 0 °), the vertical direction in the ROI (in the figure). Among the pixels arranged in the Y-axis direction), the number of pixels having a luminance of less than a threshold (for example, 200) (that is, dark pixels having a luminance of less than or equal to the threshold) is counted for each X coordinate of the ROI (step S200). Here, each time the CPU 11 detects one pixel whose brightness is equal to or less than the threshold value at a certain X coordinate, the CPU 11 increases the number of pixels at the X coordinate by one.

Further, the CPU 11 stores the number of counted pixels in the count data shown in FIG. 7. The count data includes the number of pixels whose brightness is equal to or less than the threshold among the pixels arranged in the vertical direction (Y-axis direction in the figure) in one ROI corresponding to each of the plurality of rotation angles and each of the plurality of thresholds. The count value corresponding to the number of the pixels is the data recorded for each X coordinate of the ROI.

Next, as a function of the counting means 33, the CPU 11 weights the number of pixels counted in step S200 using a weighting coefficient corresponding to the threshold value (for example, 200) (step S202). Here, the weighting coefficient may be recorded in the count data in a state associated with each of the plurality of threshold values, for example. Further, the weighting coefficient may be set so that the number of pixels is heavily weighted as the threshold value becomes smaller. The CPU 11 weights the number of pixels for each X coordinate counted using a predetermined threshold value using a weighting coefficient corresponding to the threshold value, and records it in the count data as a count value corresponding to the number of the pixels.

Next, the CPU 11 determines whether or not the number of pixels has been counted for all the threshold values (step S204), and when the count processing has been performed for all the threshold values (step S204: YES), the process of step S206 is performed. Transition. If the count processing is not performed for all the threshold values in step S204 (step S204: NO), the process proceeds to the process of step S200 using the threshold values for which the count processing is not performed.

Further, the CPU 11 determines whether or not the number of pixels has been counted for all rotation angles (step S206), and when the count processing has been performed for all rotation angles (step S206: YES), in step S210. Move to processing. If the counting process is not performed for all the threshold values in step S206 (step S206: NO), the ROI is rotated by the rotation angle at which the counting process is not performed (step S208), and the process is performed in step S200. Transition.

In the process of step S208, the CPU 11 is rotated by a predetermined rotation angle around the origin of the ROI (in the example of the figure, the upper left end portion), as shown in FIGS. 8 (a) to 8 (b), for example. .. Then, the CPU 11 performs the process of step S200 for the ROI after rotation shown in FIG. 8 (b).

As a function of the detecting means 34, the CPU 11 extracts the X coordinate which is the boundary of the product based on the number of counted pixels when the counting process is performed for all the rotation angles (step S206: YES) (step S210). ). For example, the CPU 11 counts the number of pixels when using any threshold value in a state where the ROI is rotated at any rotation angle among a plurality of X coordinates in the ROI, or a count value corresponding to the number of pixels. X-coordinates of a predetermined value (for example, the size (number of pixels in the vertical direction) in the vertical direction (Y-axis direction in the figure) of the ROI x the predetermined specified value (for example, 0.0 to 1.0)) or more of the product It may be extracted as a boundary.

Further, the CPU 11 extracts, as a product boundary, X coordinates in which the total number of pixels or count values for each threshold value corresponding to at least one rotation angle is equal to or greater than a predetermined value among the plurality of X coordinates in the ROI. You may.

When there are a plurality of X-axis coordinates that satisfy a predetermined condition regarding the number of pixels within a predetermined range (for example, 100 pixels), the CPU 11 extracts the coordinates of any one of them as the boundary of the product. You may. Here, the predetermined condition regarding the number of pixels is, for example, the number of pixels counted when the ROI is rotated at any rotation angle and any threshold value is used, or the count value corresponding to the number of pixels. May be greater than or equal to a predetermined value, or the total number of pixels or count values for each at least one threshold corresponding to at least one rotation angle may be greater than or equal to a predetermined value. Further, the CPU 11 may, for example, extract the coordinates of the X-axis having the maximum count value corresponding to the counted number of pixels or the number of pixels among the coordinates of the X-axis satisfying a predetermined condition as the boundary of the product. The CPU 11 may perform such an extraction process between, for example, step S200 and step S202, or may be performed between step S202 and step S204.

Further, when the CPU 11 extracts the X coordinate that is the boundary of the product in step S210, as shown in FIG. 9, the CPU 11 attaches the boundary line B that passes through the extracted X coordinate and extends along the Y-axis direction to the ROI. You may.

As described above, the product existing in the ROI can be detected.

(4) Main processing flow of the object detection system of the present embodiment Next, an example of the main processing flow performed by the object detection system of the present embodiment will be described with reference to the flowchart of FIG.

First, the CPU 11 of the detection device 10 receives (acquires) the image of the product shelf S transmitted from the communication terminal 20 via the communication interface unit 18 (step S300). Further, the CPU 11 stores the received image in, for example, the RAM 13 or the non-volatile memory 14.

Next, the CPU 11 of the detection device 10 identifies each of the plurality of areas in the image as an ROI by performing image processing on the acquired image (step S302). The processing content of step S302 is the same as the processing content of steps S100 to S106 of FIG. 4 described above.

Next, the CPU 11 of the detection device 10 selects one ROI from the specified plurality of ROIs (step S304), and detects the product existing in the selected ROI (step S306). The processing content of step S306 is the same as the processing content of steps S200 to S210 of FIG. 6 described above.

Further, the CPU 11 determines whether or not the product detection process has been performed for all ROIs (step S308), and ends all the processes when the detection process has been performed for all ROIs (step S308: YES). .. If the detection process is not performed for all ROIs in step S308 (step S308: NO), the process proceeds to step S304 so that the ROI for which the detection process has not been performed is selected.

As described above, according to the object detection system and the object detection method of the present embodiment, among the pixels arranged in the vertical direction (Y-axis direction) of the ROI, the pixels whose brightness is equal to or less than a predetermined threshold value (that is, the brightness is equal to or less than the threshold value). The product can be detected by extracting the coordinates that are the boundaries of the product from the coordinates in the lateral direction (X-axis direction) of the ROI based on the number of dark pixels). This makes it possible to detect the product by counting the number of pixels whose brightness is equal to or less than a predetermined threshold value. Therefore, the product in the image is compared with the case where the detection process is performed using, for example, a learning method. Can be detected by a simple process, and further, the cost required for the detection process can be suppressed.

The program for causing the computer to execute the object detection method of the present invention may be stored in a computer-readable storage medium. The storage medium on which this program is recorded may be, for example, the non-volatile memory 14 shown in FIG. Further, it may be a CD-ROM or the like that can be read by being inserted into a program reading device such as a CD-ROM drive. Further, the storage medium may be a magnetic tape, a cassette tape, a flexible disk, an MO (Magneto Optical disk) / MD (MiniDisk) / DVD (Digital Versatile Disc), or the like, or may be a semiconductor memory.

The embodiments described above are described for facilitating the understanding of the present invention, and are not described for limiting the present invention. Therefore, each element disclosed in each of the above embodiments is intended to include all design changes and equivalents belonging to the technical scope of the present invention.

For example, in the above-described embodiment, among the pixels arranged in the vertical direction (Y-axis direction) of the ROI, the laterality of the ROI is based on the number of pixels whose brightness is equal to or less than a predetermined threshold value (that is, dark pixels whose brightness is equal to or less than the threshold value). The case of extracting the coordinates that are the boundaries of the product from the coordinates in the direction (X-axis direction) has been described as an example, but among the pixels arranged in the horizontal direction (X-axis direction) of the ROI, the pixels whose brightness is equal to or less than a predetermined threshold value. Based on the number of, the coordinates that are the boundaries of the products may be extracted from the coordinates in the vertical direction (Y-axis direction) of the ROI.

Further, in the above-described embodiment, the case where the image of the product shelf S is transmitted to the detection device 10 using the communication terminal 20 has been described as an example, but the present invention is not limited to this case. For example, when the image of the product shelf S is stored in a storage medium readable by a computer (for example, a USB (Universal Serial Bus) memory or the like), the detection device 10 accesses the storage medium to obtain the product. The image of the shelf S may be acquired.

Further, in the above-described embodiment, the detection device 10 is configured to realize the functions of the acquisition means 31, the specific means 32, the counting means 33, and the detection means 34, but the configuration is not limited to this. For example, at least a part of these means may be realized by the communication terminal 20.

According to the object detection system and object detection method of the present invention as described above, a predetermined object in an image can be detected by a simple process, and for example, the display status of a product in a physical store can be confirmed based on image data. Since it can be suitably used for a system or the like, its industrial applicability is extremely high.

10 ... Detection device 20 ... Communication terminal 31 ... Acquisition means 32 ... Specific means 33 ... Counting means 34 ... Detection means B ... Boundary line S ... Product shelf

Claims (6)

An acquisition means for acquiring an image of a predetermined object, and
When the horizontal direction of the acquired image is the X-axis direction and the vertical direction of the image is the Y-axis direction, among the pixels arranged in one of the X-axis and the Y-axis, the pixels whose brightness is equal to or less than a predetermined threshold. A counting means that counts a number for each coordinate of the other of the X-axis and the Y-axis,
A detection means for detecting the predetermined object by extracting the coordinates of the other, which is the boundary of the predetermined object in the image, based on the number of counted pixels.
Equipped with
When a plurality of the other coordinates satisfying a predetermined condition regarding the number of pixels are present in a predetermined range, the detection means extracts any one of the other coordinates as a boundary of the predetermined object. , Object detection system. An acquisition means for acquiring an image of a predetermined object, and
When the horizontal direction of the acquired image is the X-axis direction and the vertical direction of the image is the Y-axis direction, among the pixels arranged in one of the X-axis and the Y-axis, the pixels whose brightness is equal to or less than a predetermined threshold. A counting means that counts a number for each coordinate of the other of the X-axis and the Y-axis,
A detection means for detecting the predetermined object by extracting the coordinates of the other, which is the boundary of the predetermined object in the image, based on the number of counted pixels.
Equipped with
The counting means counts the number of the pixels using each of the plurality of threshold values when a plurality of the threshold values are provided, and counts using any one of the plurality of threshold values. An object detection system that weights the number of pixels according to the value of any of the above thresholds. The object detection system according to claim 1 or 2 , wherein the counting means counts the number of the pixels in a state where the acquired image is rotated by a predetermined rotation angle. When the predetermined object exists in each of the plurality of areas in the image, the image is provided with a specific means for identifying each of the plurality of areas by performing image processing on the image.
The counting means counts the number of pixels in at least one of the plurality of areas.
The object detection system according to any one of claims 1 to 3 , wherein the detection means detects the predetermined object existing in at least one area. The computer
Steps to acquire an image of a given object, and
When the horizontal direction of the acquired image is the X-axis direction and the vertical direction of the image is the Y-axis direction, among the pixels arranged in one of the X-axis and the Y-axis, the pixels whose brightness is equal to or less than a predetermined threshold . And the step of counting the number of each for each coordinate of the other of the X-axis and the Y-axis,
A step of detecting the predetermined object by extracting the coordinates of the other, which is the boundary of the predetermined object in the image, based on the number of counted pixels.
Perform each step of
In the step of detecting the predetermined object, when a plurality of the other coordinates satisfying the predetermined condition regarding the number of pixels are present within the predetermined range, any one of the other coordinates is used as the predetermined object. Object detection method to extract as a boundary . The computer
Steps to acquire an image of a given object,
When the horizontal direction of the acquired image is the X-axis direction and the vertical direction of the image is the Y-axis direction, among the pixels arranged in one of the X-axis and the Y-axis, the pixels whose brightness is equal to or less than a predetermined threshold. A step of counting a number for each coordinate of the other of the X-axis and the Y-axis,
A step of detecting the predetermined object by extracting the coordinates of the other, which is the boundary of the predetermined object in the image, based on the number of counted pixels.
Perform each step of
In the counting step, when a plurality of the threshold values are provided, the number of the pixels is counted using each of the plurality of threshold values, and the counting is performed using any one of the plurality of threshold values. An object detection method in which the number of pixels is weighted according to the value of any of the threshold values.

Images