JP4249967B2 - Object detection method and object detection apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a monitoring device using an imaging device, and more particularly to an object detection method and an object detection device that automatically detect an object that has entered the imaging field of view from a video signal input from the imaging device. .
[0002]
[Prior art]
2. Description of the Related Art Video monitoring apparatuses using an imaging device such as a television camera (hereinafter referred to as a TV camera) have been widely used. In a monitoring system using such a video monitoring device, a monitoring person detects an image acquired by capturing an image of the inside of the monitoring field of view of an intruding object such as a person or a car that enters the monitoring field of view, There was a way to do it while looking at the display screen of the monitor. However, from such a manned monitoring method, a system that automatically detects an intruding object by performing image processing on an image input from an image input means of a TV camera and obtains a predetermined notification or alarm treatment. Is becoming required.
[0003]
In order to realize such a system, first, an intruding object in the field of view is detected by a difference method or the like. The difference method is a method in which an input image obtained from a TV camera is compared with a reference background image created in advance, that is, an image in which an object to be detected is not captured, and a difference in luminance value is obtained for each pixel. Is detected as an object. An example of these processes will be described with reference to FIGS. 6, 7, and 8. FIG.
[0004]
FIG. 6 is a diagram for explaining the flow of processing for detecting an object that has entered the field of view of the imaging apparatus by the difference method. S01 is the input image, S02 is the reference background image, S03 is the difference image between the input image and the reference background image, S04 is the binary image of the difference image, S05 is the difference processing unit, S06 is the binarization processing unit, and S07 is the input Humanoid object shown in image S01, S08 is humanoid difference image in difference image S03 corresponding to humanoid object S07, S09 is human in binary image S04 corresponding to humanoid difference image S08 Represents a type object (a human type binarized image).
[0005]
In FIG. 6, first, an input image S01 of 320 × 240 pixels, for example, is input from the camera. Next, in the difference processing unit S05, a difference between pixel values for each pixel between the input image S01 and the reference background image S02 created in advance is calculated, and a difference image S03 is obtained.
At this time, the humanoid object S07 in the input image S01 appears as a humanoid difference image S08 in the difference image S03. Then, in the binarization processing unit S06, the value of the pixel whose difference value is less than a predetermined threshold value (for example, less than 20) is set to “0” for each pixel of the difference image S03, and the pixel value equal to or greater than the threshold value Is replaced with “255” (this example will be described with an example in which one pixel is treated as 8 bits) to obtain a binary image S04. At this time, the humanoid object S07 captured in the input image S01 is detected as a humanoid object S09 in the binarized image S04. The above is the outline of the object detection process by the difference method (for example, refer to Patent Document 1).
[0006]
FIG. 7 is a diagram for explaining a problem that occurs in the object detection method using the difference method described with reference to FIG. 6. M01 is an input image, M01a is a region of an intruding object, and M01b and M01c are regions.
FIG. 7 shows a case where there are plants and the like in addition to the intruding object in the input image M01. If the difference method is applied in such a case, in addition to the area M01a of the intruding object, areas such as vegetation swaying by the wind, for example, the area M01b and the area M01c are detected. Therefore, it is determined by the difference method that there are three intruding objects due to the regions M01a, M01b, and M01c.
[0007]
Conventionally, a mask process as shown in FIG. 8 has been widely used for such a problem. FIG. 8 is a diagram for explaining an example of a conventional solution to the problem of the difference method described with reference to FIG. M02 is the input image, M02a is the area of the intruding object, M02b, M02c
Is an area.
In FIG. 8, a mask area M02b is set for an area where plants and the like exist. The mask area M02b is processed as a dead zone that prevents an intruding object from being detected (mask processing), or a threshold value in the mask area is set high so that vegetation swaying in the wind is not erroneously detected. In this way, it is determined that an intruding object exists only in the area M02a in the input image M02.
[0008]
The conventional mask region prevents, for example, a false detection caused by shaking of a plant or the like in a scene where the plant or the like exists in the monitoring visual field. However, if an intruding object to be detected enters the mask area, it becomes a dead zone and is not detected after being masked, or the threshold value becomes high and it is difficult to detect. Therefore, when the wind in the monitoring area stops and the shaking of the vegetation stops, it is necessary to cancel the unnecessary mask area to prevent the detection of an intruding object (missing detection).
In the conventional object detection apparatus, when the mask area is canceled, it is determined by the operator. However, in this method, the operator needs skill, and the operator has to appropriately determine whether or not the mask area needs to be released, and there is a problem that the burden on the operator is large.
[0009]
[Patent Document 1]
Japanese Patent Laid-Open No. 7-79429 (page 4-7, FIG. 1)
[0010]
[Problems to be solved by the invention]
In the above-described conventional technique, in order to accurately detect only an intruding object within the monitoring visual field, a difference method is used, and mask processing is used for the purpose of preventing vegetation shaking and reflection of light from a puddle. Also, when the wind etc. weakens and the area to be masked becomes small, it is necessary to minimize the oversight of the object to be detected caused by the mask process in order to perform the mask process properly. The operator has to cancel the unnecessary mask area in accordance with the state of the monitoring environment, which has the disadvantage that the operator is burdened and the operator needs to be skilled.
In addition, when releasing the mask area, it is preferable to release the mask areas in order from the highest priority to be released, but there is a problem that it is burdensome for the operator to determine the order and skill is required. .
The object of the present invention is to eliminate the above-mentioned drawbacks, facilitate or automate manual operation requiring skill, and reduce the burden on the operator so that an appropriate mask area can be set. Another object of the present invention is to provide a highly reliable object detection method and apparatus.
[0011]
[Means for Solving the Problems]
In order to achieve the above object, an object detection method of the present invention is an object detection method for detecting an object by a difference method from an image acquired by an imaging apparatus, and sets a predetermined region of the image as a mask region, In the object detection method in which the set mask area is excluded from the object detection target, the mask area is numbered as a partial mask area for each new or added area, and the partial mask area is based on the assigned number. The setting is canceled every time.
Also preferably, the cancellation of the setting for each mask area in the object detection method of the present invention is performed in the reverse order of the numbered order.
[0012]
That is, the object detection method of the present invention includes an object detection step for detecting an object in an image signal sequentially input from an imaging device, a mask processing step for masking a detection object with a predetermined mask region, and a partial region of the mask region. In the object detection method including a mask area release step for excluding a partial mask area that is a mask processing target, the mask area is at least one partial mask area numbered based on an order in which the partial mask areas are set In other words, an object that has entered the imaging field of view of the imaging apparatus is detected.
Preferably, the mask area releasing step of the object detection method of the present invention is characterized in that the numbered partial mask areas are released based on the number.
Further preferably, the mask area releasing step of the object detection method of the present invention is characterized in that the partial mask areas numbered in the reverse order of the numbers are released.
[0013]
Preferably, the object detection method of the present invention further includes operation input means for inputting an operator operation, and the mask area releasing step is performed on the partial mask areas numbered based on the operator's operation input from the operation input means. Among them, at least one is released.
Preferably, the mask area release step of the object detection method of the present invention includes a mask area automatic release determination step for determining whether or not to release the partial mask area from the mask area based on the object detection result of the object detection step. In addition, at least one of the partial mask areas is automatically canceled.
Preferably, the mask region automatic release determination step of the object detection method of the present invention performs release determination of the partial mask region having the largest number among the numbered partial mask regions.
Preferably, the mask region automatic release determination step of the object detection method of the present invention performs release determination for all of the numbered partial mask regions.
Preferably, the mask area releasing step of the object detection method of the present invention releases all the numbered partial mask areas having a number equal to or greater than the number of the numbered partial mask areas.
[0014]
The object detection apparatus of the present invention is an object detection method for detecting an object in an image signal sequentially input from an imaging apparatus, masking the detected object with a predetermined mask area, and detecting an object in an imaging field of view. An imaging device that sequentially captures a monitoring visual field range, an image input interface that sequentially converts a video signal acquired by the imaging device into an image signal, an image processing unit that processes an image signal converted by the image input interface, and an operator operation Operation input means for inputting a predetermined image signal and output means for outputting a predetermined image signal. The image processing means numbers the partial mask areas, which are partial areas of the mask area stored in the image memory means, in the set order. Determine whether to cancel the partial mask area based on the detection result, and cancel the operation input from the operation input means or the partial mask area An object that has entered the imaging field of view of the imaging device is detected by releasing the partial mask area from the mask area based on one of the determination results as to whether or not it is a new mask area. .
[0015]
DETAILED DESCRIPTION OF THE INVENTION
In the object detection method of the present invention, in order to solve the conventional problem that skill is required for canceling the mask area and the problem that the burden on the operator is large, the mask area is divided into partial areas (hereinafter referred to as mask areas). The partial area is referred to as a partial mask area) and numbered and managed, and the mask area can be easily maintained in an appropriate state by manually or automatically releasing the partial mask area based on the number. it can.
[0016]
FIG. 5 is a block diagram showing an embodiment of the hardware configuration of the object detection apparatus according to an embodiment of the present invention. E01 is a TV camera, E02 is a zoom lens, E03 is an electric swivel base (hereinafter referred to as camera pan head), E04 is an operation means, E04a is a first button attached to the operation means E04, and E04b is attached to an operation means E04. E04c is a direction key attached to the operation means E04, E05a is an image input I / F (I / F: Interface), E05b is a pan head control I / F, E05c is a lens control I / F, E05d Is the operation input I / F, E05e is the image memory, E05f is the image output I / F, E05g is the alarm output I / F, E05h is the CPU (CPU: Central Processing Unit), E05i is the program memory, E05j is the work memory, E05k Is a data bus, E05 is an image input I / F E05a, pan head control I / F E05b, lens control I / F E05c, operation input I / F E05d, image memory E05e, image output I / F E05f, alarm output I / F E05g, CPU E05h, program memory E05i, work memory E05j, and data bus E05k Intruding object monitoring apparatus, E06 is output monitor, E07 is a warning light.
[0017]
The camera E01 is coupled to the image input I / F E05a, the zoom lens E02 is coupled to the lens control I / F E05c, the camera head E03 is coupled to the pan head control I / F E05b, and the operation means E04 is the operation input I. / F E05d, output monitor E06 is connected to image output I / F E05f, and warning light E07 is connected to alarm output I / F E05g. Image input I / F E05a, pan head control I / F E05b, lens control I / F E05c, input I / F E05d, image memory E05e, image output I / F E05f, alarm output I / F E05g, CPU E05h The program memory E05i and the work memory E05j are coupled to the data bus E05k.
[0018]
In FIG. 5, a TV camera E01 equipped with a zoom lens E02 mounted on a camera head picks up an image of a monitoring target (field-of-view range). The captured video signal is stored in the image memory E05e from the image input I / F E05a through the data bus E05k. The CPU E05h analyzes the image stored in the image memory E05e in the work memory E05j according to the program stored in the program memory E05i. The CPU E05h controls the zoom lens E02 from the data bus E05k via the lens control I / F E05c or the camera head E03 via the head control I / F E05b according to the processing result. Change the field of view of the camera E01, turn on the warning lamp E07 via the alarm output I / F E05g, and display, for example, the intruding object detection result image on the monitor monitor E06 via the image output I / F E05f. The image memory E05e also includes a template image holding device for storing registered template images. All the embodiments described below will be described using the hardware configuration of the object detection apparatus described with reference to FIG.
[0019]
A first embodiment of the present invention will be described with reference to FIGS. FIG. 2 is a flowchart for explaining a processing process according to an embodiment of the present invention.
In the embodiment of FIG. 2, the masked areas are released in the reverse order to the masked areas set by the operator's predetermined operation on the masked areas assigned at the time of setting the masked areas. Is.
[0020]
First, in an initialization step 101, initialization is performed on an image memory for image storage secured in the image memory E05e, variables stored in the work memory E05j, and the like used in the object detection process.
Next, in the image input step 102, for example, an input image S01 having a width of 320 pixels, a height of 240 pixels, and 8 bits / pixel is acquired from the camera E01.
In the difference processing step 103, the difference for each pixel between the input image S01 obtained in the image input step 102 and the reference background image S02 recorded in advance in the image memory E05e is calculated to obtain a difference image S03.
[0021]
Next, in the binarization processing step 104, threshold processing is performed for each pixel of the difference image S03 obtained in the difference processing step 103. If the pixel value is less than the threshold (for example, 20), the pixel value Is equal to “0”, and when it is equal to or greater than the threshold value, the pixel value is set to “255” to obtain a binary image S04.
Next, the mask processing step 105 performs the mask processing described with reference to FIG.
In the mask processing step 105, the binarized image S04 obtained in the binarization processing step 104 is masked based on the mask area stored in the image memory E05e. In the above example, the mask area is set to “255” as the pixel value of the pixel in the non-masked area as the mask image. Furthermore, in the binarization processing step 104, the binarized image sets “255” for pixels above the threshold value and “0” for pixels below the threshold value for each pixel of the difference image. That is, in the process of removing the mask area from the binarized image, the binarized image is “255” (that is, the difference between the input image and the reference background image is equal to or greater than a predetermined threshold), and the mask image is “255”. Only the pixels of “(that is, the region not to be masked) are extracted.
By this process, the detection area within the mask area (the pixel having the pixel value “255” in the binarized image) can be excluded from the object detection process.
[0022]
Next, in the operator setting operation presence / absence determination step 106, it is determined whether or not the operator has pressed, for example, the first button E04a of the operation means E04. If there is an operator setting operation, the process branches to the mask area setting processing step 107. If there is no operator operation, the process branches to the operator release operation presence / absence determination step 201.
In the mask area setting processing step 107, for example, a pixel having a value of “255” binarized in the binarization processing step 104 is additionally set as a new mask area in the mask area stored in the image memory E05e. Update the mask area.
[0023]
Next, in mask area numbering step 108, the mask areas are numbered and stored in the form of an image. For example, as shown in FIG. 10, the pixel value of the pixel in the area not set as the mask area is set to “255”, and the mask set first for each area added to the mask area in the mask area setting processing step 107 The pixel value of the pixel in the area is numbered and stored as “1”, the pixel value of the pixel in the second set mask area is “2”, and the like.
[0024]
Next, in the operator release operation presence / absence determination step 201, it is determined whether or not the operator has pressed, for example, the second button E04b of the operation means E04.
If there is an operator release operation, the process branches to the maximum number mask area release step 202, and if there is no operator operation, the process branches to the object presence determination step 116.
In the maximum number mask area releasing step 202, one mask area having the largest number is released. For example, in the image 804 shown in FIG. 10 (d), in the image memory E05e, the mask area with the largest number is the area 804d and the number is “4”. Therefore, as shown in FIG. 10C, the mask area is canceled by returning the pixel value of the pixel of the mask area 804d with the maximum number “4” to “255”.
[0025]
Next, in the object presence determination step 116, the “S255” cluster S09 is detected from the masked binarized images obtained in the binarization processing step 104 and the mask processing step 105 by, for example, a labeling method. It is determined whether or not an object exists. If it exists, the process branches to an alarm / monitor output step 115. If not, the process returns to the image input step 102.
In the alarm / monitor display step 115, for example, an image of the target object is displayed on the output monitor E06, or the warning lamp E07 is turned on to notify that an intruding object exists.
[0026]
An example of the effect of canceling the mask area based on the number in this way will be described with reference to FIG. FIG. 9 is a diagram showing an example of a mask area to be set that changes with time.
In FIG. 9, a mask region such as a mask region 801a (FIG. 9 (a)) is set at time T1, and the wind becomes stronger at time T2, so that false detections increase and operator operations Then, a new mask area 802b (FIG. 9B) is additionally set. In addition, it can be judged from the fact that the wind has become strong due to an increase in false detections.
Similarly, assume that a mask region 803c (FIG. 9C) is additionally set at time T3, and a mask region 804d (FIG. 9D) is additionally set at time T4.
That is, in this example, the mask area is set stepwise according to the wind intensity.
[0027]
At this time, it is assumed that the wind has weakened and it is necessary to reduce the mask area. Since the range of false detection differs depending on the wind intensity, the mask area 804d, which was necessary when the wind was strong, is not necessary, and a mask area like the necessary mask area 803 at time T3 should be set. It is.
When the wind further weakens, the mask area 803c becomes unnecessary, and a mask area such as the necessary mask area 802 should be set. In other words, the necessary mask area changes in the reverse order of the setting.
[0028]
With respect to the strength of the wind, as shown in the table of FIG. 11, the number of necessary masks increases as the wind becomes stronger, and the number of necessary masks decreases as the wind becomes weaker.
Therefore, as in this embodiment, the number “1” is assigned to the mask area 801a set at time T1, the number “2” is assigned to the mask area 802b newly set at time T2, and the like. A number “3” is assigned to the mask area 803c and a number “4” is assigned to the mask area 804d.
That is, the mask area additionally set when the wind is strong is given a larger number than the mask area set when the wind is weaker. Then, by releasing in reverse order from number 4 which is the maximum number when releasing the mask area, the mask area can be easily and appropriately not lost when the mask is released by the operator. Can be released.
[0029]
A second embodiment of the present invention will be described with reference to FIGS. 3, 11 and 12. FIG. FIG. 3 is a flowchart for explaining a processing process according to an embodiment of the present invention.
The embodiment shown in FIG. 3 is configured such that a plurality of areas can be released at once from the mask area of the largest order in the first embodiment of the present invention.
That is, in this embodiment, a mask area release candidate display step 301 is added to the first embodiment of the present invention, and the mask area release step 202 having the largest number is added to the mask area selection step 302 to be released and the numbered mask. This is replaced with the area release step 114. The other steps are the same as those in the first embodiment described with reference to FIG.
[0030]
In FIG. 3, in a mask area cancellation candidate display step 301, in order to select a mask area to be canceled by the operator, display is performed as a candidate that can cancel the numbered and set mask areas.
For example, for the mask area as shown in FIG. 12 (b) recorded in the image memory E05e with respect to FIG. 12 (a) as the input image, an outline is drawn for each different mask area and assigned to each mask area. For example, the number is displayed as a number. In the mask area selection step 302 to be released, a mask area to be released is selected by an operator's operation using the operation means E04, and a number assigned to the mask area, for example, a white numeral shown in FIG. In this way, the mask area cancellation candidates are displayed.
[0031]
In the mask area selection step 302 to be released, a mask area to be released is selected by an operator's operation using the operation means E04 (for example, a number is selected or an area is selected), and a number assigned to the mask number is selected. Take out. This operator's operation may specify the mask area number with a numerical value, or use a pointing device such as a mouse to specify the mask release candidate displayed by GUI operation or the displayed number. May be.
[0032]
Next, in a numbered mask area releasing step 114, all mask areas numbered more than the number of the selected mask area are released.
For example, in the mask area cancellation candidate shown in FIG. 12 (c), if it is specified to release the mask area of the number “3” (corresponding to the mask number “3”), the wind intensity shown in FIG. On the other hand, as can be seen by referring to the table of necessary mask areas, the mask area corresponding to the number “3” should also be released because the mask area corresponding to the number “4” (corresponding to the mask number “4”) should also be released. At the same time, the mask area corresponding to the number “4” is automatically canceled.
[0033]
As described above, one of the numbered mask areas is designated by the operator, and the mask area numbered higher than the number of the mask area is canceled, and the mask area and the mask which the operator has decided to release are released. A mask area having a higher priority to be released than the area (in the example of FIG. 9, a mask area required in a windy scene) can be released at the same time. Therefore, the mask area can be released easily and appropriately.
[0034]
A third embodiment of the present invention will be described with reference to FIG. FIG. 4 is a flowchart for explaining a processing process according to an embodiment of the present invention.
In the embodiment of FIG. 4, in the first embodiment of the present invention, the mask area released by the operator's operation can be automatically determined and released. That is, in this embodiment, in the first embodiment of the present invention, the operator release operation presence / absence determination step 201 is replaced with a mask area detection presence / absence determination step 110, an area release timer reset step 111, an area release timer check step 112, This is replaced with the predetermined time elapsed determination step 113. The other steps are the same as those in the first embodiment described with reference to FIG.
[0035]
In FIG. 4, in-mask area detection presence / absence determination step 110 is performed based on the numbered mask areas stored in the work memory E05j and the numbered masks based on the binarized image stored in the image memory E05e. It is determined whether or not a detection area (a pixel having a pixel value “255” in the binarized image) exists in the area.
In this embodiment, the presence / absence of a detection area in the mask area having the largest number is determined.
If there is a detection area in the numbered mask area, the process branches to an area release timer reset step 111. If no detection area exists, the process branches to an area release timer check step 112.
[0036]
Next, in the area release timer reset step 111, the area release timer in the mask area is reset.
In the area release timer check step 112, an elapsed time since the area release timer was last reset is calculated.
In this embodiment, this processing is performed by the timer function of the CPU E05h. However, if the timer of the CPU E05h cannot be used, an external timer may be used.
[0037]
Next, in the predetermined time elapsed determination step 113, the elapsed time is a predetermined time (for example, 10 minutes) based on the elapsed time calculated after the area cancellation timer is reset calculated in the area cancellation timer check step 112. If it has elapsed, the process branches to the mask area release step 202 having the largest number, and if the predetermined time has not elapsed, the process branches to the object existence determination step 116.
[0038]
As described above, the detection area in the mask area with the largest number is determined, and the mask area with the largest number is released based on the elapsed time since the last detection area was detected. It can be automatically canceled in the reverse order of setting the area.
As a result, the mask area can be automatically released without the operator knowing that the wind has weakened. Accordingly, it is possible to automatically set an appropriate mask area and to accurately detect an intruding object.
[0039]
A fourth embodiment of the present invention will be described with reference to FIG. FIG. 1 is a flowchart illustrating a processing process according to an embodiment of the present invention.
In the third embodiment of the present invention, in the third embodiment of the present invention, a plurality of unnecessary mask areas can be released at a time from the one that has been released one by one from the largest unnecessary mask areas. Is.
That is, in this embodiment, a check completion judging step 109 and a calculation number counter incrementing step 115 are added to the third embodiment of the present invention, and the mask area releasing step 202 having the largest number is added to the numbered mask area releasing step 114. It has been replaced with. The other steps are the same as those in the third embodiment described with reference to FIG.
[0040]
In the embodiment of FIG. 1, the mask area is canceled based on the elapsed time since the last detection in the detection area for all numbered mask areas. Accordingly, in the check completion determination step 107, it is determined whether the release determination processing has been completed for all mask areas from the value of the calculation number counter stored in the work memory E05j and the total number of mask areas numbered. Then, if it is completed, the process branches to the object presence determination step 116, and if it is not completed, the process branches to the detection in / out of mask area step 110.
[0041]
Next, in the numbered mask area releasing step 114, the mask area of the calculated number designated by the calculation number counter and the mask area numbered larger than the calculated number are all released.
Next, in the calculation number counter increment step 115, the calculation number counter stored in the work memory E05j is incremented, and the process returns to the check completion determination step 109. That is, the mask area is released based on the elapsed time since the last detection in the detection area for the next mask area.
[0042]
As described above, for all numbered mask areas, the mask area is released based on the elapsed time since the last detection in the detection area, and the mask area to be released is a mask for which no object is detected for a predetermined time. All mask areas with numbers greater than the area and its number are released. As a result, for example, an unnecessary mask area can be automatically released simultaneously based on the number even under conditions where the wind suddenly stops blowing.
Accordingly, it is possible to automatically set an appropriate mask area and to accurately detect an intruding object.
[0043]
【The invention's effect】
As described above, according to the present invention, according to the present invention, the mask area is canceled or designated in the reverse order to the mask setting based on the number assigned at the time of mask area setting for each mask area. By simultaneously canceling the mask areas having the numbers greater than or equal to the number, the mask areas can be released appropriately and with less burden on the operator. Therefore, the manual operation requiring skill can be facilitated or automated, the burden on the operator can be reduced and an appropriate mask area can be set, and a highly reliable object detection method and apparatus have been realized.
[Brief description of the drawings]
FIG. 1 is a flowchart illustrating a processing flow according to an embodiment of the present invention.
FIG. 2 is a flowchart illustrating the flow of processing according to an embodiment of the present invention.
FIG. 3 is a flowchart illustrating a processing flow according to an embodiment of the present invention.
FIG. 4 is a flowchart illustrating a processing flow according to an embodiment of the present invention.
FIG. 5 is a block diagram showing a configuration of a video monitoring apparatus according to an embodiment of the present invention.
FIG. 6 is a diagram for explaining the operation of object detection processing by a conventional difference method.
FIG. 7 is a diagram for explaining a region of a detected object detected by a conventional difference method.
FIG. 8 is a diagram for explaining the operation of conventional mask processing.
FIG. 9 is a diagram for explaining the operation of conventional mask processing.
FIG. 10 is a diagram showing an example of a change in a mask area to be set that changes with time.
FIG. 11 is a table showing necessary mask areas with respect to wind intensity according to an embodiment of the present invention.
FIG. 12 is a diagram for explaining an embodiment of the present invention.
[Explanation of symbols]
101: initialization step, 102: image input step, 103: difference processing step, 104: binarization processing step, 105: mask processing step, 106: operator setting operation presence / absence determination step, 107: mask region setting processing step, 108 : Mask area numbering step, 109: Check completion determination step, 110: Detection presence / absence determination step in the mask area, 111: Area release timer reset step, 112: Area release timer check step, 113: Predetermined elapsed time determination step 114: Numbered mask area release step, 115: Calculation number counter increment step, 116: Object presence determination step, 117: Alarm / monitor display step, E01: TV camera, E02: Zoom lens, E03: Camera head, E04 : Operation means, E04a: first button, E04b: second button, E04c: direction key, E05: intruding object monitoring device, E05a: Image input I / F, E05b: Head control I / F, E05c: Lens control I / F, E05d: Operation input I / F, E05e: Image memory, E05f: Image output I / F, E05g: Alarm output I / F, E05h: CPU, E05i: Program memory, E05j: Work memory, E05k: Data bus, E06: Output monitor, E07: Warning light, M01: Input image, M01a, M01b, M01c: Area, M02: Input image M02a, M02, M02c: area, S01: input image, S02: reference background image, S03: difference image, S04: binarized image, S05: difference processing unit, S06: binarization processing unit, S07: humanoid S08: Humanoid difference image, S09: Humanoid binary image.

Claims (4)

An object detection method for detecting an object in an image obtained from an imaging apparatus, wherein a predetermined area of the image is set as a mask area, and the set mask area is excluded from an object detection target.
The mask region is numbered every area to be new or additional, unset as a mask area of the numbered pre Symbol areas, one for the numbered were the reverse order object detection method according to claim Rukoto done more or once. The object detection method according to claim 1,
Release settings as a mask area of the region, object detection wherein a row cracking Turkey when the predetermined time object in the mask area is not detected. An object detection apparatus for detecting an object in an image obtained from an imaging apparatus, wherein the predetermined area of the image is set as a mask area, and the set mask area is excluded from an object detection target.
The mask region is numbered every area to be new or additional, unset as a mask area of the numbered pre Symbol areas, one for the numbered were the reverse order or object detection apparatus according to claim Rukoto done more at once. The object detection device according to claim 3 ,
Release settings as a mask area of the region, the object detecting device comprising a row cracking Turkey when the predetermined time object in the mask area is not detected.

Images