JP2014143551A - Image processing apparatus, imaging apparatus, and image processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing apparatus capable of detecting an object without having an object detection disabled period when changing an imaging range.SOLUTION: An image processing apparatus 106 includes: background information preparation means for preparing background information from a reference video; and object detection means for detecting object information in an input video from the difference between the background information and the input video. When an imaging range is changed, the background information preparation means re-prepares new background information from the reference video generated after changing the imaging range, and prepares provisional background information corresponding to a common area among the reference videos generated before and after changing the imaging range by means of the background information before the change which was prepared before changing the imaging range and stored in storage means 110. In terms of the detection of object information in the input video generated after changing the imaging range, the object detection means employs the difference between the provisional background information and the input image for the above difference until the preparation of the new background information is completed.

Description

  The present invention relates to an image processing technique for detecting object information in a video generated by imaging using a so-called background difference method.

  Background subtraction as an image processing technique for detecting in real time object information (hereinafter also simply referred to as an object) indicating the position and size of an area where an object other than the background exists in an image generated by an imaging device such as a surveillance camera The law is known. For example, Patent Document 1 discloses a method of accurately extracting an object from a video using a background difference method.

  However, in the background subtraction method, a background image (background information) generated in a state where the imaging range, which is the range of the object field imaged by the imaging device, is fixed, and generated by imaging an object entering the background The object is detected from the difference from the input image. For this reason, if the imaging range at the time of object detection is greatly changed with respect to the imaging range when the background image is generated by zooming, panning and tilting of the imaging device, the already created background image becomes meaningless, It is necessary to create new background information. In addition, object detection (that is, monitoring) cannot be performed during a period necessary to recreate new background information.

  In order to avoid the occurrence of an object non-detectable period due to the re-creation of the background information, Patent Document 2 previously sets a change destination of an imaging range by zooming, panning, and tilting, and corresponds to the imaging range of the change destination. A method for storing background information to be stored in a memory is disclosed.

JP 2000-020722 A JP 2002-051329 A

  However, in the method disclosed in Patent Document 2, it is necessary to recreate background information when zooming, panning, and tilting are performed in an imaging range other than a predetermined change destination. Will occur.

  The present invention provides an image processing apparatus, an imaging apparatus, and an image processing program capable of performing object detection with almost no object detection impossible period even when the imaging range is arbitrarily changed.

  An image processing apparatus according to an aspect of the present invention performs processing on a video generated by an imaging apparatus that captures an object scene within an imaging range. The image processing device detects background information from the reference image generated by the imaging device, and background information generating means, and detects object information in the input video from the difference between the background information and the input video generated by the imaging device. An object detecting means for storing and a storage means for storing background information. And when the imaging range is changed, the background information creation means recreates new background information from the reference video generated after the change, and among the reference videos generated before and after the change of the imaging range Provisional background information corresponding to the common area is created using the pre-change background information created before the change of the imaging range and stored in the storage unit. Further, the object detection means detects the difference between the provisional background information and the input image as the difference until the creation of new background information is completed in the detection of the object information in the input video generated after the change of the imaging range. It is characterized by using.

  Note that an image pickup apparatus including an image pickup unit that picks up an object scene in an image pickup range, an image pickup range change unit that changes the image pickup range, and the image processing apparatus also constitute another aspect of the present invention.

  An image processing program according to another aspect of the present invention is a computer program that causes a computer to execute processing on an image generated by an imaging device that captures an object scene within an imaging range. The image processing program stores, in a computer, background information creation processing for creating background information from a reference video generated by the imaging device, and an object in the input video based on a difference between the background information and the input video generated by the imaging device. An object detection process for detecting information and a process for storing background information in the storage means are performed. Then, the image processing program causes the computer to re-create new background information from the reference video generated after the change when the imaging range is changed in the background information creation process, and to change the imaging range. Provisional background information corresponding to the common area among the reference images generated before and after is created using the pre-change background information created before the change of the imaging range and stored in the storage means. Further, in the object detection process for the input video generated after the imaging range is changed, the object information is detected using the difference between the provisional background information and the input image as the difference until the creation of new background information is completed. It is characterized by that.

According to the present invention, even when the imaging range is arbitrarily changed, the object information can be detected with almost no period during which the object information cannot be detected.

1 is a block diagram showing a configuration of a surveillance camera that is Embodiment 1 of the present invention. FIG. 5 is a flowchart illustrating processing performed by the monitoring camera according to the first embodiment. The flowchart which shows the process performed by the surveillance camera which is Example 2 of this invention. 10 is a flowchart illustrating processing performed by a monitoring camera that is Embodiment 3 of the present invention. The figure which shows the change of the imaging range (video) by zoom-in, zoom-out, and panning / tilting. The figure which shows the object information detected in the surveillance camera of each Example.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 shows the configuration of an imaging apparatus that is an embodiment of the present invention. Here, the monitoring camera 100 will be described as an imaging device. However, the following configuration can be applied to various imaging apparatuses such as a video camera other than the surveillance camera and a digital still camera having a moving image imaging function.

  In FIG. 1, reference numeral 101 denotes a lens unit. The lens unit 101 includes an imaging lens (imaging optical system) including a movable lens such as a variable power lens or a focus lens, and an actuator that moves the movable lens or rotates the lens unit 101 in a panning direction or a tilting direction. Have Reference numeral 102 denotes a lens control unit which controls the operation of the lens unit 101 in accordance with an instruction from the CPU 105. Specifically, an image pickup range that is a range of an object imaged by the camera 100 by moving the zoom lens to perform zooming or rotating the lens unit 101 to perform panning or tilting. To control.

  Reference numeral 103 denotes an image pickup device as an image pickup means for picking up an image of a subject formed by the image pickup lens (photoelectric conversion), and is constituted by a photoelectric conversion device such as a CCD sensor or a CMOS sensor. An imaging control unit 104 converts an analog imaging signal output from the imaging element 103 into digital data.

  As a main controller, the CPU 105 performs arithmetic processing based on control of each unit in the camera 100 and information obtained from each unit. Reference numeral 106 denotes a video processing unit as an image processing apparatus, which performs processing for creating background information (background information creation processing) and processing for detecting object information (object detection processing) described later. A recording medium control unit 107 controls writing and reading of data to and from a synchronous dynamic random access memory (hereinafter referred to as SDRAM) 108 as a recording medium. Reference numeral 109 denotes an SDRAM control unit that controls writing and reading of data to and from the SDRAM 110 serving as a storage unit in accordance with an instruction from the CPU 105. When the imaging control unit 104 converts the imaging signal into digital data, the SDRAM 110 is used as a primary storage unit. At this time, data is written to and read from the SDRAM 110 under the control of the SDRAM control unit 109. The converted digital data is written to the recording medium 108 via the recording medium control unit 107. The SDRAM 110 can also temporarily store the internal state of the camera 100, video data waiting for processing, data generated in background image creation processing and object detection processing, and the like.

  The flowchart of FIG. 2 illustrates processing performed by the CPU 105 and the video processing unit 106 in association with zooming (zoom-in) to the telephoto side in the camera 100 of the present embodiment. The CPU 105 and the video processing unit 106 execute the following processing according to a control program that is a computer program held in the recording medium 108 as a computer mounted on the camera. The CPU 105 functions as an imaging range changing unit together with the lens control unit 101, and the video processing unit 106 functions as a background information creation unit and an object detection unit.

  In step S201, the CPU 105 starts zoom processing to the telephoto side (hereinafter referred to as zoom-in processing) in response to a zooming instruction from the outside to the telephoto side. When the zoom-in process is completed in step S202, the CPU 105 proceeds to step S203.

  In step S <b> 203, the video processing unit 106 starts a process for recreating background information (new background information) in response to the imaging range being changed by the zoom-in process. At this time, the video processing unit 106 keeps the background information before zoom-in processing (background information before change) stored in the SDRAM 110 without overwriting, and secures a new area on the SDRAM 110 to create a new imaging range. Start creating corresponding new background information.

  Here, the background information is information indicating the luminance and color of the background in the video (reference video) generated by imaging in a state where no object exists in the background within the imaging range. In the background subtraction method employed in the object detection processing in the present embodiment, the input video is calculated from the difference between the background information and a video (input video) generated by imaging in a state where an object can exist within the imaging range. Object information indicating the position and size of the area in which the object exists is detected.

  In this embodiment, the average value obtained by adding the pixel values of the video data obtained from the image sensor 103 for a predetermined number of frames for each pixel is used as the background information. In general, when background information is created from a video generated by imaging, time for a plurality of frames is required until the creation of stable background information is completed. That is, while the background information re-creation process is being performed in step S203, the object detection process by the background subtraction method using the background information in the process of creation cannot be performed. For this reason, in this embodiment, the characteristic processing described below is performed.

  In step S204, the video processing unit 106 performs a process of extracting a common area that is a common area among videos (reference videos) generated before and after the change of the imaging range by the zoom-in process. An image corresponding to the imaging range before the zoom-in process is shown on the left side of FIG. 5A, and an image corresponding to the imaging range after the zoom-in process is shown on the right side. The video after the zoom-in process is an enlarged part of the video before the zoom-in process. For this reason, a common area | region with the image | video after a zoom-in process can be extracted from the image | video before a zoom-in process.

Here, as a method of extracting a common area from two videos generated before and after the change of the imaging range by the zoom process, there is a method of estimating using a change rate of the imaging range. For example, a region indicated by a dotted line in the left diagram of FIG. 5A indicates a common region with the video illustrated in the right diagram. In the common area, if the number of pixels of the image in the left figure is 1280 × 1024 pixels and the change rate of the imaging range is 20% (however, in the horizontal direction)
It is estimated that (1280 × 1024 pixels) × (horizontal 100−20%) = 1024 × 768 pixels.

  Next, in step S205, the video processing unit 106 selects a partial background corresponding to the common area obtained in step S204 from the pre-change background information stored in the SDRAM 110 in the previous step S203. Information is extracted as provisional background information.

  Next, from step S206, the video processing unit 106 performs object detection processing by the background subtraction method using the background information extracted in step S205. First, in step S206, the video processing unit 106 determines whether or not new background information re-creation processing corresponding to the imaging range after the zoom-in processing started in step S203 has been completed. If it is determined that the re-creation process has not yet been completed and is continuing, the process proceeds to step S207.

  In step S <b> 207, the video processing unit 106 acquires an input video (video generated by imaging after zoom-in processing) that is an object of object detection processing, and develops the input video data on the SDRAM 110.

  Next, in step S208, the video processing unit 106 performs a resizing process that matches the number of pixels of the acquired input video with the number of pixels of the provisional background information extracted in step S205. Specifically, a process of compressing the input video by a general algorithm such as an average pixel method so that the number of pixels matches the number of pixels in the common area calculated in step S204 (the number of pixels in the provisional background information). I do.

  In step S209, the video processing unit 106 executes background difference processing as object detection processing between the provisional background information extracted in step S205 and the input video resized in step S208. Thereby, the video processing unit 106 detects the difference between the provisional background information and the input video as object information.

  Next, in step S210, the video processing unit 106 determines whether object information has been acquired by the background difference process in step S209. FIG. 6 shows object information (data structure) including identification ID, coordinates and size information in the input image, as an example of object information. When the object information is acquired, the video processing unit 106 proceeds to step S211 and starts a process (moving object / unmoving object detection process) for analyzing the existence of the moving object / unmoving object using the object information. If the object information has not been acquired (detected), the process returns to step S206.

  In the moving object / non-moving object analysis processing in step S211, the video processing unit 106 uses an object information identification ID and coordinate information as moving object determination, and there is an object moving over a plurality of frames of the input video. It is determined whether or not. In addition, the image processing unit 106 uses the identification ID, coordinate information, and a timer function as an unmoving object determination, and uses an object (non-moving object) that has been stopped for a certain period of time at a certain position over a plurality of frames of the input image. It is determined whether or not (object) exists.

  Next, in step S212, the video processing unit 106 determines whether the moving object / unmoving object analysis process in step S211 is being executed. If it is being executed, the video processing unit 106 proceeds to step S213, sets the moving object / unmoving object analysis process execution flag, and then returns to step S206. On the other hand, if the moving object / unmoving object analysis process has not been executed, the video processing unit 106 proceeds to step S214 to cancel the moving object / unmoving object analysis process execution flag, and then returns to step S206.

  The series of processing from step S206 to step S214 is repeated with the time for a certain frame as an upper limit until it is determined in the determination processing in step S206 that the re-creation processing of new background information has been completed. If the video processing unit 106 determines in step S206 that the re-creation process of new background information has been completed, the video processing unit 106 proceeds to step S215.

  In step S215, the video processing unit 106 determines whether or not the moving object / unmoving object analysis process execution flag set and canceled in steps S213 and S214 is set. When the moving object / unmoving object analysis process execution flag is set (moving object / unmoving object analysis process is being executed), the video processing unit 106 performs step S207 to prioritize the moving object / unmoving object analysis process. Return to.

  On the other hand, when the moving object / unmoving object analysis process execution flag is canceled (the moving object / unmoving object analysis process is not being executed), the video processing unit 106 proceeds to step S216. In step S216, the video processing unit 106 starts object detection processing based on background difference processing using new background information having a sufficient amount of information originally necessary as background information. That is, while the moving object / non-moving object analysis process is being executed, even if the re-creation process of the new background information is completed, the moving object / non-moving object analysis is not immediately performed without moving to the object detection process using the new background information. After completing the process, the process proceeds to an object detection process using new background information.

  As described above, in the present embodiment, even during the process of creating new background information required by the zoom-in process, the pre-change background information created before the zoom-in process is used for the input image after the zoom-in process. Object detection processing can be performed. Thereby, even when the imaging range is arbitrarily changed by the zoom-in process, the object information can be detected with almost no period during which the object information cannot be detected.

  In the present embodiment, the case has been described where the input video resizing process is performed in step S208 in order to perform the object detection process using the provisional background information extracted in step S205. However, instead of the resizing process of the input video, it is also possible to create the temporary background information by performing a resizing process that increases the number of pixels of the provisional background information so as to match the number of pixels of the input picture.

  Next, Example 2 of the present invention will be described. The flowchart of FIG. 3 shows processing performed by the CPU 105 and the video processing unit 106 in association with zooming (zoom out) to the wide angle side.

  In the first embodiment that describes the case where the zoom-in process is performed, the imaging range after the zoom-in process is included as a part of the imaging range before the zoom-in process. Therefore, provisional background information is extracted as part of background information before zoom-in processing (pre-change background information), and resize processing is performed on the input video generated by imaging after zoom-in processing.

  In contrast, in this embodiment in which zoom-out processing is performed, the imaging range before zoom-out processing is included as part of the imaging range after zoom-out processing. For this reason, the background information before zoom-out processing (background information before change) is resized to create provisional background information.

  The CPU 105 and the video processing unit 106 execute the following processing according to a control program that is a computer program held in the recording medium 108 as a computer mounted on the camera.

  In step S301, the CPU 105 starts zoom-out processing in response to an external zooming instruction to the wide angle side. When the zoom-out process is completed in step S302, the CPU 105 proceeds to step S303.

  In step S <b> 303, the video processing unit 106 starts a process for recreating background information (new background information) in response to the change of the imaging range by the zoom-out process. At this time, the video processing unit 106 keeps the background information before zoom processing (pre-change background information) stored in the SDRAM 110 without overwriting, and secures a new area on the SDRAM 110 to create a new imaging range. Start creating corresponding new background information.

  Next, in step S304, the video processing unit 106 performs a process of extracting a common area from videos (reference videos) generated before and after the change of the imaging range by the zoom-out process. The left side of FIG. 5B shows an image corresponding to the imaging range before the zoom-out process, and the right side shows an image corresponding to the imaging range after the zoom-out process. The video after the zoom-out process is a video with a wider imaging range including the entire video before the zoom-out process. For this reason, it is possible to extract a common area with the video before zoom-out processing from the video after zoom-out processing.

A region indicated by a dotted line in the right diagram of FIG. 5B indicates a common region with the video illustrated in the left diagram. In the common area, the number of pixels of the video in the right figure is 1280 × 1024 pixels, and the change rate of the imaging range is 20% (however, in the horizontal direction)
It is estimated that (1280 × 1024 pixels) × (horizontal 100−20%) = 1024 × 768 pixels.

  Next, in step S305, the video processing unit 106 matches the number of pixels of the background information before the change stored in the SDRAM 110 in the previous step S303 with the number of pixels in the common area obtained in step S304. Resize processing to reduce to Specifically, a process of compressing the background information before change so that the number of pixels matches the number of pixels in the common area calculated in step S304 is performed.

  Next, in step S306, the video processing unit 106 determines whether the new background information re-creation process corresponding to the imaging range after the zoom-out process started in step S303 has been completed. If it is determined that the re-creation process has not yet been completed and is continuing, the process proceeds to step S307.

  In step S <b> 307, the video processing unit 106 acquires an input video (video generated by imaging after zoom-out processing) to be subjected to object detection processing, and develops the input video data on the SDRAM 110.

  Next, in step S308, the video processing unit 106 extracts a video portion corresponding to the common area (hereinafter referred to as an input partial video) from the acquired input video.

  In step S309, the video processing unit 106 performs background difference processing as object detection processing between the provisional background information resized in step S305 and the input partial video extracted in step S308. Thereby, the video processing unit 106 detects the difference between the provisional background information and the input partial video as object information.

  The processing after step S310 is the same as the processing after step S210 in the first embodiment.

  As described above, even in the present embodiment, even during the process of creating new background information required by the zoom-out process, the pre-change background information created before the zoom-out process is used. Object detection processing can be performed on the input image. Thereby, even when the imaging range is arbitrarily changed by the zoom-out process, the object information can be detected with almost no period during which the object information cannot be detected.

  In the present embodiment, a case has been described in which the portion corresponding to the common area is extracted from the input video in step S308 in order to perform the object detection process using the temporary background information resized in step S305. However, instead of the resize process of the provisional background information, it is also possible to perform a resize process that increases the number of pixels of the input video so as to match the number of pixels of the pre-change background information (provisional background information).

  Next, Embodiment 3 of the present invention will be described. The flowchart of FIG. 4 illustrates processing performed by the CPU 105 and the video processing unit 106 in association with panning (rotation of the imaging range in the horizontal direction) and tilting (rotation of the imaging range in the vertical direction). .

  In this embodiment in which panning / tilting processing is performed, unlike in Embodiments 1 and 2 in which zoom-in / zoom-out processing is performed, an image in one imaging range includes the entire image in the other imaging range before and after the processing. There is no relationship. For this reason, in this embodiment, the video processing unit 106 performs processing to determine whether or not a common area exists in the video before and after panning / tilting processing (hereinafter referred to as pan / tilt processing). FIG. 5C shows a change in the imaging range when pan / tilt processing is performed and a common area in the imaging range before and after the change. When there is such a common area, a process of extracting from the pre-change background information created and held before the pan / tilt process is performed.

  The CPU 105 and the video processing unit 106 execute the following processing according to a control program that is a computer program held in the recording medium 108 as a computer mounted on the camera.

  In step S401, the CPU 105 starts pan / tilt processing in response to an external panning or tilting instruction. When the pan / tilt process is completed in step S402, the CPU 105 proceeds to step S403.

  In step S403, the video processing unit 106 starts a process of recreating background information (new background information) in response to the change of the imaging range by the pan / tilt process. At this time, the video processing unit 106 keeps the background information before zoom processing (pre-change background information) stored in the SDRAM 110 without overwriting, and secures a new area on the SDRAM 110 to create a new imaging range. Start creating corresponding new background information.

  Next, in step S404, the video processing unit 106 determines whether there is a common area in the video (reference video) generated before and after the change of the imaging range by the pan / tilt process. If the common area exists, the process proceeds to step S405, and if not, the process proceeds to step S415.

  In step S405, the video processing unit 106 extracts a portion corresponding to the common area from the pre-change background information stored in the SDRAM 110 in the previous step S403, and uses this as provisional background information.

  Next, in step S406, the video processing unit 106 determines whether the new background information re-creation process corresponding to the imaging range after the pan / tilt process started in step S403 has been completed. If it is determined that the re-creation process has not yet been completed and is continuing, the process proceeds to step S407.

  In step S <b> 407, the video processing unit 106 acquires an input video (video generated by imaging after pan / tilt processing) to be subjected to object detection processing, and develops the input video data on the SDRAM 110.

  Next, in step S408, the video processing unit 106 extracts a video portion corresponding to a common area (hereinafter referred to as an input partial video) from the acquired input video.

  In step S409, the video processing unit 106 executes background difference processing as object detection processing between the provisional background information extracted in step S405 and the input partial video extracted in step S408. Thereby, the video processing unit 106 detects the difference between the provisional background information and the input partial video as object information.

  The processing after step S410 is the same as the processing after step S410 in the first embodiment. In step S415, the video processing unit 106 performs the same determination as in step S406, waits for the completion of the process for recreating new background information, and then proceeds to step S417 (same as step S216 in the first embodiment).

  As described above, even in this embodiment, even during the process of creating new background information required by the pan / tilt process, the pan / tilt is performed using the pre-change background information created before the pan / tilt process. Object detection processing can be performed on the input image after processing. Thereby, even when the imaging range is arbitrarily changed by the pan / tilt process, the object information can be detected with almost no period during which the object information cannot be detected.

  In the present embodiment, a case has been described in which, when it is determined that there is no common area in the images before and after the pan / tilt process, the process waits until the new background information re-creation process is completed. However, the change of the imaging range by the pan / tilt process is limited so that the common area always exists and the size of the common area does not become smaller than a predetermined size capable of the object detection process (that is, panning or The operating range of tilting may be limited). In this case, the CPU 105 functions as an imaging range change limiting unit.

  Furthermore, an operation mode for managing the validity / invalidity of such pan / tilt process restrictions may be prepared.

  In each of the above-described embodiments, the case where the image processing unit 106 as the image processing apparatus is included in the imaging apparatus has been described. However, the image processing apparatus may be configured by a personal computer different from the imaging apparatus. In this case, the image processing apparatus acquires an image (video signal) from the imaging apparatus by wired or wireless communication or via a storage medium such as a semiconductor memory or an optical disk. In this case, the image processing program may cause the personal computer to execute the processing performed by the video processing unit 106 in the flowcharts illustrated in FIGS.

  Each embodiment described above is only a representative example, and various modifications and changes can be made to each embodiment in carrying out the present invention.

  It is possible to provide an imaging apparatus and an image processing apparatus that can perform good object detection processing even when the imaging range is changed.

100 surveillance camera 105 CPU
106 Video processing unit 110 SDRAM

Claims (9)

An image processing device that performs processing on a video generated by an imaging device that captures an object scene within an imaging range,
Background information creating means for creating background information from a reference image generated by the imaging device;
Object detection means for detecting object information in the input video from the difference between the background information and the input video generated by the imaging device;
Storage means for storing the background information,
The background information creation means recreates new background information from the reference video generated after the change when the imaging range is changed, and the reference video generated before and after the change of the imaging range. Provisional background information corresponding to the common area is created using the pre-change background information created before the change of the imaging range and stored in the storage means,
In the detection of the object information in the input video generated after the change of the imaging range, the object detection means detects the provisional background information and the input image as the difference until the creation of the new background information is completed. An image processing apparatus using a difference between the two. When the imaging range is changed by zooming in the imaging device,
The object detection unit obtains the difference after performing a resizing process for adjusting the number of pixels of the provisional background information and the input video generated after the imaging range is changed. The image processing apparatus described.   When zooming in is performed as the zooming, the object detection unit reduces the number of pixels of the input video generated after the change of the imaging range to match the number of pixels of the provisional background information as the resizing process. The image processing apparatus according to claim 2, wherein processing or processing for increasing the number of pixels of the provisional background information to match the number of pixels of the input video generated after changing the imaging range is performed.   When zooming out is performed as the zooming, the object detection unit reduces the number of pixels of the provisional background information to the number of pixels of the input video generated after the change of the imaging range as the resizing process. The image processing apparatus according to claim 2, wherein the matching process or the process of increasing the number of pixels of the input video generated after changing the imaging range is performed to match the number of pixels of the provisional background information. When the imaging range is changed by panning or tilting in the imaging device,
The background information creating means determines whether or not a common area exists in the imaging range before and after the change, and creates the provisional background information when the common area exists. The image processing apparatus according to 1. In the case where the imaging range is changed by panning or tilting in the imaging device,
2. The image processing apparatus according to claim 1, further comprising an imaging range change limiting unit that limits the change of the imaging range so that the size of the common area does not become smaller than a predetermined size. . The object detection means includes
A moving object / unmoving object detection process for detecting a moving object and an unmoving object as the object in the input image;
Even if the creation of the new background information is completed before the moving object / unmoving object detection process is completed, the detection of the object information using the new background information is started after the moving object / unmoving object detection process is completed. The image processing apparatus according to claim 1, wherein the image processing apparatus is an image processing apparatus. An imaging means for imaging an object scene within an imaging range;
Imaging range changing means for changing the imaging range;
An imaging apparatus comprising: the image processing apparatus according to claim 1. An image processing program as a computer program that causes a computer to execute processing on an image generated by an imaging device that images an object scene within an imaging range,
In the computer,
Background information creation processing for creating background information from the reference video generated by the imaging device;
Object detection processing for detecting object information in the input video from the difference between the background information and the input video generated by the imaging device;
Processing to save the background information in a storage means,
In the background information creation process, when the imaging range is changed, new background information is created again from the reference video generated after the change, and the reference generated before and after the change of the imaging range is performed. Provisional background information corresponding to a common area of the video is created using the pre-change background information created before the change of the imaging range and stored in the storage means,
In the object detection process for the input video generated after the change of the imaging range, the object is obtained using the difference between the provisional background information and the input image as the difference until the creation of the new background information is completed. An image processing program for detecting information.