JP2009122990A - Image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that when a moving object is detected by comparing a photographic image with a prepared photographic image in an area other than a mask area where the moving object is not detected, the moving object cannot be detected, which is located within a predetermined mask area for a moving range of a blocking object. <P>SOLUTION: A moving area is detected by moving area detecting means from an image obtained by image recording means which records both the moving object and the blocking object. The blocking object is detected by blocking object detecting means in accordance with operation storing means for storing operation information in advance such as the moving range, the moving direction or the speed of the blocking object. The area of the detected blocking object is decided as the mask area and an area excluding the mask area is set as a moving object detecting area by moving object detecting area setting means. The moving object is detected by moving object detecting means from the moving object detecting area and the moving area detected by the moving area detecting means. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing technique for detecting a moving object from captured image data.

  As a system for detecting a moving object from photographed image data, for example, there is a system disclosed in Japanese Patent Application Laid-Open No. 2004-068515. In the system according to Patent Document 1, a region where a moving object is not detected is designated in advance as a mask region, and a region other than this region is compared with a captured image at a different time. Thus, a moving object is detected.

JP 2004-068515 A

  The conventional system for detecting moving objects is configured as described above, and it is necessary to designate in advance as an area where the moving object is not detected, for example, the area where the occluded object moves, as the area where the occluded object moves. When all are masked, the moving object cannot be detected when the moving object is within the moving range of the shielding object.

  The present invention has been made to solve the above-described problems, and by using the motion information of the shielding object, it is determined whether or not the shielding object is to be masked, and only the portion blocked by the shielding object. The object is to detect a moving object by masking.

  An image processing apparatus according to the present invention includes an image capturing unit that captures both a moving object and a shielding object that performs a specific operation different from the moving object, and a moving region from an image obtained by the image capturing unit. A moving area detecting means for detecting movement information, a movement information storage means for storing movement information such as a moving range, a moving direction, and a speed of the shielding object in advance, and a shielding object from the image based on the storage information of the movement information storage means Occluding object detecting means for detecting the object, determination of whether or not the area of the occluding object obtained by the occluding object detecting means is a mask area, and a moving object for setting a range excluding the mask area as a moving object detecting area Detection area setting means, and a moving object detection means for detecting a moving object from the area set by the moving object detection area setting means and the moving area detected by the moving area detection means .

  According to the image processing apparatus of the present invention, since the moving object detection area setting unit sets the area to be masked according to the movement detection result of the shielding object in the shielding object detection unit, only when the shielding object moves. The moving range can be masked, and a moving object can be detected with high accuracy.

Embodiment 1 FIG.
1 is a block diagram showing an image processing apparatus according to Embodiment 1 of the present invention. In the figure, an image capturing unit 1 captures an image of a moving object and a shielding object that performs a specific operation different from the moving object using an imaging device such as the same camera.
The movement area detection unit 2 detects an area having movement from the image obtained by the image photographing unit 1 as a movement area.
The motion information storage unit 3 is a storage device that holds in advance the moving range, moving direction, speed, and the like of an object that is a shielding object.
The shielding object detection unit 4 performs processing for monitoring the detection and movement of the shielding object in the image captured by the image capturing unit 1 based on the moving range, moving direction, and speed of the shielding object in the motion information storage unit 3.
The moving object detection region setting unit 5 determines whether or not the shielding object region obtained by the shielding object detection unit 4 is a mask region in the image captured by the image capturing unit 1 and the range excluding the mask region is a moving object. Processing to set as a detection area is performed.
The moving object detection unit 6 executes processing for detecting a moving object from the moving object detection region set by the moving object detection region setting unit 5 and the moving region detected by the moving region detection unit 2.
In the first embodiment, a vehicle is assumed as a moving object, and a bar that permits or rejects passage of the vehicle as a shielding object is assumed.

FIG. 2 is a flowchart showing the processing contents of the image processing apparatus according to Embodiment 1 of the present invention.
First, the image photographing unit 1 photographs a moving object and a shielding object that performs a specific operation different from the moving object with the same camera or the like (step ST11). In this embodiment, it is assumed that the image of FIG. 3 is taken. Here, the bar is used as a shielding object, and the vehicle is detected as a moving object.

The movement area detection unit 2 detects an area having a movement from the image obtained from the image photographing unit 1 as a movement area (step ST12). Here, as a method of detecting the moving region, detection is performed using a method based on a difference between the image obtained by the image photographing unit 1 and the background image shown in FIG. For example, as a method for detecting a moving region based on a difference from a background image, a method described in Japanese Patent Application Laid-Open No. 07-336694 is used.
FIG. 5 is an image example of the moving area obtained by the difference between the captured image of FIG. 3 and the background image of FIG. 4 by the moving area detecting unit 2, and the area shown in black means the moving area. Here, the method based on the difference from the background image is used as the method for detecting the moving region, but other methods may be used.
The moving area detection unit 2 determines whether or not a moving area is included on the image. If there is no moving area on the image, the process returns to step ST11. If there is a moving area, the process of step ST14 is performed. Proceed (step ST13). Here, since the moving area is included, the process proceeds to step ST14.

The shielding object detection unit 4 sets the initial state previously stored in the motion information storage unit 3 as the shielding object monitoring area. FIG. 6 shows an example of information held in the operation information storage unit 3. In the figure, the coordinate points A0 (x, y) to D0 (x, y) and the moving range AT (x, y) to FT (x, y), shielding for determining the range to monitor the initial state of the shielding object The moving direction (dx, dy) of the object and the speed (Vx, Vy) of the shielding object are shown.
FIG. 7 shows an example in which the shielding object detection unit 4 acquires the initial state from the motion information storage unit 3 and sets the shielding object monitoring region. In the figure, the inside of the broken line indicates the shielding object monitoring area.
Next, when the change in the pixel value in the shielding object monitoring area exceeds a predetermined threshold, the shielding object detection unit 4 detects that the shielding object has started to move from the initial state (step ST14). Here, as a method of detecting the movement of the shielding object, the detection is performed based on the change value of the pixel in the shielding object monitoring area, but other methods may be used.

The moving object detection area setting unit 5 determines whether or not the shielding object is in an initial state (step ST15). If the shielding object is in the initial state, the process proceeds to step ST17 using the shielding object monitoring area as the mask area, and if not, the process proceeds to step ST16. Here, as shown in FIG. 3, since the bar which is a shielding object starts moving from the initial state and moves, the process proceeds to step 16.
Next, the moving object detection region setting unit 5 acquires the moving range of the shielding object from the motion information storage unit 3, and sets this as a mask region (step ST16). FIG. 8 shows an example in which the moving range of the shielding object is a mask area. In the drawing, a black portion represents a mask area.
The moving object detection area setting unit 5 sets a range excluding the mask area from the movement area obtained by the movement area detection unit 2 as a moving object detection area (step ST17). FIG. 9 shows an example of a moving object detection area obtained by removing the mask area in FIG. 8 from the moving area in FIG.

The moving object detection unit 6 detects a rectangle including them as a moving object from the moving object detection region set by the moving object detection region setting unit 5 and the moving region detected by the moving region detection unit 2 (step ST18). . FIG. 10 shows a rectangular example of the moving object detected by the moving object detection unit 6. Here, as a method for detecting a moving object, a rectangle that includes a region is detected as a moving object. However, after removing noise by expansion / contraction processing or the like, it is disclosed in Japanese Patent Laid-Open No. 2000-242787. You may make it perform using the method of detecting a mobile body detection using a template, and another method.
The processes in steps 11 to 18 are repeatedly executed to continuously detect the moving object.

  As described above, according to the first embodiment, since the moving object detection region setting unit 5 sets the area to be masked according to the movement detection result of the shielding object in the shielding object detection unit 4, the shielding object is set. This has the effect of being able to mask the moving range only when is moved.

In this embodiment, the moving object detection region setting unit 5 obtains the moving range of the shielding object from the motion information storage unit 3 and uses this as the mask region, but the shielding object stored in the motion information storage unit 3 As shown in FIG. 11, the range in which the shielding object moves is represented by a plurality of pieces of position information and the moving direction and speed of the position information, and the position information is represented by a count value. Then, the initial state is used as a reference, and the count in the initial state is 0.
In step 15, when the position of the shielding object is not in the initial state, counting is performed, the position of the shielding object corresponding to each count is obtained from the motion information storage unit 3 as shown in FIG. You may perform so that it may set as an area | region.

  FIG. 12 shows an example in which the mask area of the shielding object at the corresponding count is obtained and the range excluding this is set as the moving object detection area.

  FIG. 13 shows an example in which a moving object is detected by the moving object detection unit 6 from the moving object detection region of FIG.

  With the configuration as described above, according to this embodiment, there is an effect that a moving object can be detected with high accuracy by masking only a portion blocked by the shielding object.

Embodiment 2. FIG.
Next, an image processing apparatus according to another embodiment of the present invention will be described. The configuration diagram showing the second embodiment of the present invention is FIG. 1 as in the first embodiment.
Also in the second embodiment, a vehicle is assumed as a moving object, and a bar that permits or rejects passage of the vehicle as a shielding object.
FIG. 14 is a flowchart showing the processing contents of the image processing according to the second embodiment of the present invention. In the figure, processing units that perform the same processing as in the first embodiment are denoted by the same reference numerals, and description thereof is omitted.

Similar to the operation of the first embodiment, the processing of step 11 to step 13 is performed on the image captured by the image capturing unit 1 of the image processing apparatus.
The moving area detection unit 2 divides the image obtained from the image capturing unit 1 into partial images (step ST201). In this embodiment, it is assumed that the moving area detection unit 2 divides M partial images into a total of M × N partial images, M in the horizontal direction and N in the vertical direction. FIG. 15 is an example in which the image of FIG. 3 is divided into partial images. In this drawing, the j-th partial image in the horizontal direction and the i-th vertical direction are expressed as Bij (1 ≦ j ≦ M, 1 ≦ i ≦ N).

  Next, the moving area detection unit 2 detects the motion vector of each divided partial image using a matching method as disclosed in Japanese Patent Laid-Open No. 5-75914 (step ST202). FIG. 16 shows an example of motion vectors detected using the image of FIG. 15 and the frame image at the previous time. Here, the motion vector in the partial image Bij of the t-th frame is expressed as Ft_ij (1 ≦ j ≦ M, 1 ≦ i ≦ N, 1 ≦ t ≦ L, L is the total number of images captured by the image capturing unit). . In this embodiment, an image is divided into partial images, and motion vectors are detected by a method as disclosed in Japanese Patent Laid-Open No. 5-75914. However, “Lucas BD and Kanade T” “An iterative image registration technique with an A method of detecting a motion vector in pixel units as disclosed in “application to stereo vision.” Proceedings of Imaging understanding workshop, pp 121-130, 1981 ”may be used.

  The shielding object detection unit 4 collates the partial image where the shielding object shown in FIG. 17 and the motion vector of the partial image shown in FIG. Then, a partial image having the same motion vector is detected as a shielding object (step ST203). Here, the partial image where the shielding object is located on the image is denoted as RBij, and the motion vector of the partial image RBij in the t-th frame is denoted as RBt_ij (1 ≦ j ≦ M, 1 ≦ i ≦ N, 1 ≦ t ≦ L , L is the total number of images captured by the image capturing unit).

The moving object detection area setting unit 5 determines whether to mask a partial image having a motion vector determined as a shielding object based on the result of the shielding object detection unit 4 (step ST204). If the partial image having the motion vector corresponding to the shielding object is not detected in the shielding object detection unit 4, the process proceeds to step ST17. If a partial image having the same motion vector is detected, the process proceeds to step ST16.
In step ST16 and step 17, detailed processing is omitted because the same processing as in the first embodiment is performed. However, the moving object detection unit 6 includes a range and a moving region excluding the mask region of the moving object detection region setting unit 5. A moving object is detected from the area detected by the detection unit 2.

FIG. 18 shows an example of a moving object detection region obtained by masking a partial image having a motion vector corresponding to a shielding object from the moving region of FIG.
FIG. 19 shows an example in which the moving object detection unit 6 detects a moving object from the area set by the moving object detection area setting unit 5 shown in FIG.

  As described above, according to the second embodiment, the motion vector obtained by the moving area detection unit 2 and the motion vector of the occlusion object previously stored in the motion information storage unit 5 are collated by the occlusion object detection unit 4. Then, by setting a partial image having a similar motion vector as a mask region and detecting the moving object by the moving object detection unit 6, it is possible to mask only the position of the shielding object, and the moving object becomes the shielding object. Even when there is an overlapping positional relationship, an effect of detecting a moving object is achieved.

  In this embodiment, the shielding object detection unit 4 collates the motion vector of the shielding object previously held in the motion information storage unit 5 with the motion vector of the partial image obtained by the movement area detection unit 2. Although the method has been described by way of example, it may be executed by changing or adding to a process of detecting, as a shielding object, a region in which the motion vectors of partial images at different independent positions have a symmetrical motion vector at the time of collation. .

  The image processing apparatus according to the present invention is applied to a system that can detect a moving object such as a car with certainty, and takes and monitors a railroad crossing, a parking lot, or a vehicle entrance / exit of a toll road with a camera.

1 is a configuration diagram illustrating an image processing apparatus according to Embodiment 1 of the present invention; 4 is a flowchart illustrating processing contents of the image processing apparatus according to the first embodiment. 6 is an explanatory diagram of a captured image example of the image capturing unit according to Embodiment 1. FIG. 3 is a background image photographed in advance and stored in an operation information storage unit according to Embodiment 1. FIG. 6 is an explanatory diagram of an example of a moving region image detected according to Embodiment 1. FIG. 6 is an explanatory diagram of an example of information held by an operation information storage unit according to Embodiment 1. FIG. 6 is an explanatory diagram of a setting example of a shielding object monitoring area according to Embodiment 1. FIG. 6 is an explanatory diagram of an example of a mask region according to the first embodiment. FIG. 6 is an explanatory diagram of an example of a moving object detection region according to Embodiment 1. FIG. 6 is an explanatory diagram of a moving object example detected by a moving object detection unit according to Embodiment 1. FIG. It is explanatory drawing of the other example of information which the operation | movement information storage part by Embodiment 1 is doing. It is explanatory drawing of the example of a moving object detection area | region by the information of FIG. It is explanatory drawing of the example of a moving object detected from the moving object detection area | region of FIG. It is a flowchart which shows the processing content of the image processing apparatus by Embodiment 2 of invention. It is explanatory drawing of the example which divided | segmented the image by Embodiment 2 into the partial image. 10 is an explanatory diagram of an example of a motion vector of an image according to Embodiment 2. FIG. FIG. 10 is an explanatory diagram of an example of information stored in an operation information storage unit according to the second embodiment. FIG. 10 is an explanatory diagram of an example of a moving object detection area according to the second embodiment. 6 is an explanatory diagram of an example of a moving object detected according to Embodiment 2. FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1; Image pick-up part, 2; Moving area detection part, 3; Operation | movement information storage part, 4; Shielding object detection part, 5: Moving object detection area setting part, 6;

Claims (5)

  An image photographing means for photographing both a moving object and a shielding object that performs a specific operation different from the moving object, a moving area detecting means for detecting a moving area from an image obtained by the image photographing means, Motion information storage means for storing motion information such as a moving range, a moving direction, and a speed of the shield object in advance, and a shield object detection means for detecting the shield object from the image based on the storage information of the motion information storage means; Determination of whether or not the area of the shielding object obtained by the shielding object detection means is a mask area, a moving object detection area setting means for setting a range excluding the mask area as a moving object detection area, and the moving object An image processing apparatus comprising: a moving object detecting means for detecting a moving object from the area set by the detection area setting means and the moving area detected by the moving area detecting means.   The image processing apparatus according to claim 1, wherein the moving object detection region setting unit obtains a range in which the shielding object moves when the shielding object moves, and uses the range as a mask region.   The image processing apparatus according to claim 1, wherein the moving object detection region setting unit obtains the position of the shielding object from the motion information storage unit when the shielding object moves, and sets the position as a mask region.   The moving object detection area setting means divides the image obtained by the image photographing means into a plurality of partial images, and compares the motion vector of each partial image with the stored information of the motion information storage means, thereby The image processing apparatus according to claim 1, wherein the position is obtained and used as a mask region.   The image processing apparatus according to claim 1, wherein the shielding object detection unit detects a bar as the shielding object, and the moving object detection unit detects a vehicle as the moving object. .

Images