JP5821189B2 - Motion detection device and motion detection program - Google Patents

Description

  The present invention relates to a moving object detection apparatus and a moving object detection program for detecting a moving object from an image.

  In recent years, changes in weather conditions have caused frequent occurrences of rivers locally, such as guerrilla heavy rain. The Ministry of Land, Infrastructure, Transport and Tourism needs to give evacuation instructions to persons in the river's water facilities when the rivers are flooded during guerrilla heavy rain. Hydrophilic facilities are facilities such as rest facilities, water playgrounds, and promenades installed in rivers.

  Therefore, the Ministry of Land, Infrastructure, Transport and Tourism has installed CCTV (Closed Circuit Television) cameras along the rivers, and has been grasping the current state of the rivers using CCTV cameras. However, it is difficult to make a quick judgment and response when the river is flooded. Factors that make it difficult to make quick judgments and responses are the number of CCTV cameras, and because people make visual judgments, it is necessary to narrow down target areas that require quick judgments and responses. It takes time.

  The narrowing down of the target portion can be shortened by automatically detecting a person from the image, for example. As a conventional image processing for recognizing a pedestrian or a vehicle from an image, a background difference method has been conventionally known in which an input image is compared with a background image, and a pedestrian or a vehicle is recognized based on a difference in variation such as luminance. (For example, refer to Patent Document 1).

  As described above, when a person is detected from an image by the background difference method, the difference process of the amount of change is performed on the entire input image.

Japanese Patent Laid-Open No. 2002-190012

  When the difference processing of the amount of change is performed on the entire input image, disturbance (such as shaking of a plant) is detected, so that only a person is not necessarily specified and detected. Although there is a method of applying a mask to an area such as vegetation, there is no guarantee that there is no person in that area. As a result, conventionally, it has been necessary for a person to visually determine all CCTV camera images, and there is a problem that it takes time to specify whether there is a person in a hydrophilic facility.

  An object of the present embodiment is to provide a moving object detection apparatus and a moving object detection program that can quickly detect a moving object from an image.

In order to solve the above-described problem, the present embodiment is a moving object detection device that detects a moving object from an image , and stores a camera control result for each camera in response to detection of a dangerous state based on sensing data. Section for each target camera, instructing each camera to take an angle of view setting instruction to the angle of view corresponding to the maximum shooting frequency indicated by each camera control record, and the moving object detection record for each camera. by reference to a second storage unit for storing in response to the angle of view in the each target camera, preferentially moving object range of detecting a moving object most frequently indicated motion detection results corresponding to the field angle set instruction and means for the detection target range.

  In addition, what applied the arbitrary combination of the component of this embodiment, expression, or a component to a method, an apparatus, a system, a computer program, a recording medium, a data structure, etc. is also effective as an aspect of this invention.

  According to the present embodiment, it is possible to provide a moving object detection device and a moving object detection program that can quickly detect a moving object from an image.

It is a block diagram of an example of the system containing the moving body detection apparatus of a present Example. It is a block diagram of an example of DB related to rainfall station. It is a block diagram of an example of DB related to a water level station. It is a block diagram of an example of DB related to a camera station. It is a flowchart of an example showing the process sequence of the camera monitoring server. It is a flowchart of an example showing the processing procedure of the data aggregation server and the image processing server. It is a block diagram of an example of camera station PTZ information DB. It is a block diagram of an example of extraction camera station DB. It is a block diagram of an example of camera station processing range DB. It is a block diagram of an example of camera station detection point DB. It is an image figure of an example of the view angle specified by view angle information (X coordinate, Y coordinate, zoom value). It is an image figure of an example of the processing range specified by the processing range reference point (X coordinate, Y coordinate) and the processing range (X value, Y value). It is an image figure of an example in which several detection points are contained in one view angle. It is a flowchart of an example of the first camera view angle specific process. It is a flowchart of an example of a camera view angle specific process after the second time. It is a flowchart of an example of an image process. It is a flowchart of an example of a detection data update / analysis process. It is an image figure of an example of the range designation | designated process without a detection performance. It is an image figure of an example of the range designation | designated process with a detection performance. It is an image figure of an example of a detection point analysis process. It is an image figure of an example of the screen displayed on management content. It is a hardware block diagram of an example of PC.

  Next, modes for carrying out the present invention will be described based on the following embodiments with reference to the drawings. In this embodiment, a person will be described as an example of a moving object.

  FIG. 1 is a configuration diagram of an example of a system including the moving object detection device of the present embodiment. The system of FIG. 1 includes a sensor facility 1, a CCTV camera facility 2, and a river manager facility 3. The sensor facility 1 and the CCTV camera facility 2 are field facilities.

  The sensor facility 1 includes, for example, a rain gauge 11, a water level gauge 12, a flood sensor 13, and an IP node 14 installed at the site. Sensor information observed by the rain gauge 11, the water level gauge 12, and the flood sensor 13 is collected by the IP node 14 by a telemeter. The IP node 14 collects the sensor information and transmits it to the river manager facility 3 via the IP network 4.

  The CCTV camera facility 2 includes, for example, a CCTV camera 21, an encoder device 22, and an IP node 23 installed on the site. For example, more than 10,000 CCTV cameras 21 are installed around a first-class river managed by the Ministry of Land, Infrastructure, Transport and Tourism as a monitoring camera. An image captured by the CCTV camera 21 is encoded by the encoder device 22. The IP node 23 aggregates the streams of the CCTV camera 21 and transmits them to the river manager facility 3 via the IP network 4.

  The river manager facility 3 includes, for example, a river data server 31, a camera monitoring server 32, a data aggregation server 33, an image processing server 34 as an example of a moving body detection device, a PC 35 operated by the river manager, a decoder device 36, and a TV 37.

  The river data server 31 manages water level information, rainfall information, and the like observed at the site. The camera monitoring server 32 controls the CCTV camera 21 at the site. The camera monitoring server 32 manages camera information such as PTZ information that controls the angle of view of the CCTV camera 21. The PTZ information is information including P (pan): X coordinate, T (tilt): Y coordinate, and Z (zoom): zoom value.

  The data aggregation server 33 performs field data collection processing, trigger determination processing, area limitation processing, and display processing for monitoring content, which will be described later. The data aggregation server 33 includes a camera station PTZ information database (DB) 38 to be described later.

  The image processing server 34 performs camera angle-of-view specifying processing, image processing, and detection data update / analysis processing described later. The image processing server 34 includes a camera station processing range DB 39 and a camera station detection point DB 40 described later. The image processing server 34 uses the image taken by the CCTV camera 21 as an input image, performs image processing, and detects a person in a waterfront facility in the river.

  The PC 35 displays the processing result by the image processing server 34. Further, the TV 37 displays the processing result by the image processing server 34 decoded by the decoder device 36. The river manager can quickly determine the person detection location that is a target location that needs to be promptly determined and dealt with by viewing the processing result by the image processing server 34 displayed on the PC 35 and the TV 37. The PC 35 and the TV 37 are an example of monitoring content that displays the processing result by the image processing server 34. The PC 35 and the TV 37 may be processed in cooperation. 1 may omit the data aggregation server 33 and provide the image processing server 34 with the function of the data aggregation server 33.

  FIG. 2 is a configuration diagram of an example of a DB related to the rainfall station. FIG. 2A is a configuration diagram of an example of the rainfall station DB. FIG. 2B is a configuration diagram of an example of the rainfall station observation DB. The rainfall station DB has the observation station name, latitude, longitude, and threshold as data items. The rainfall station observation DB has data reception date, rainfall station location (latitude, longitude), and rainfall data as data items. The rainfall station observation DB is provided for each rainfall station. FIG. 2B shows an example of the A rainfall station. The river data server 31 stores the rain station-related DB shown in FIG. The river data server 31 manages the rainfall data observed by the rain gauge 11 at the site as shown in FIG.

  FIG. 3 is a configuration diagram of an example of a DB related to a water level station. FIG. 3A is a configuration diagram of an example of the water level station DB. FIG. 3B is a configuration diagram of an example of the water level station observation DB. The water level station DB has observation station name, latitude, longitude, and threshold as data items. The water level station observation DB has data reception date and time, water level station location (latitude, longitude), and water level data as data items. The water level station observation DB is provided for each water level station. FIG. 3B shows an example of the A water level station. The river data server 31 stores the DB related to the water level station shown in FIG. The river data server 31 manages the water level data observed by the on-site water level gauge 12 as shown in FIG.

  FIG. 4 is a configuration diagram of an example of a DB related to the camera station. FIG. 4A is a configuration diagram of an example of the camera station DB. FIG. 4B is a configuration diagram of an example of the camera station preset information DB. The camera station DB has a camera ID, a camera name, a camera station location (latitude, longitude), and a video multicast address as data items. The camera station preset information DB includes data items such as camera ID, camera name, camera station location (latitude, longitude), preset ID, control information (X coordinate, Y coordinate, zoom value), and video multicast address. The camera station preset information DB is provided for each camera station. The camera monitoring server 32 stores the DB related to the camera station shown in FIG.

  FIG. 5 is a flowchart illustrating an example of the processing procedure of the camera monitoring server. The camera monitoring server 32 performs camera control processing in step S1. The camera control process is a process in which the camera monitoring server 32 controls the CCTV camera 21 in the field. The camera monitoring server 32 can control the CCTV camera 21 to an angle of view desired by the river manager by transmitting a command including PTZ information to the CCTV camera 21 in the field. In step S <b> 2, the camera monitoring server 32 transmits PTZ information in which the angle of view of the CCTV camera 21 is controlled to the data aggregation server 33.

  FIG. 6 is a flowchart illustrating an example of processing procedures of the data aggregation server and the image processing server. In step S11, the data aggregation server 33 collects the information managed by the rainfall station related DB shown in FIG. 2 and the water level station related DB shown in FIG. The data aggregation server 33 collects information managed by the camera station-related DB shown in FIG. Further, the data aggregation server 33 receives the PTZ information from the camera monitoring server 32 and stores it in the camera station PTZ information DB 38 for each camera station.

  In step S12, the data aggregation server 33 performs a trigger determination process using the collected field data. The trigger determination process is a process that uses the collected field data to determine the presence of dangerous rainfall stations and water level stations, and if there are dangerous rainfall stations and water level stations, creates a list of dangerous rainfall stations and water level stations. It is.

  For example, when there is a rainfall station whose rainfall data exceeds the threshold or a water level station whose water level data exceeds the threshold during guerrilla heavy rain, typhoon, heavy rain, etc., the data aggregation server 33 determines that it is a dangerous rainfall station / water level station. To do. The data aggregation server 33 creates a list of dangerous rainfall stations and water level stations.

  In step S13, the data aggregation server 33 performs area limitation processing using a list of dangerous rainfall stations and water level stations. The area limitation process is a process of creating an extracted camera station DB (to be described later) by narrowing down the rain station / water level station nearby / related camera stations included in the list of dangerous rainfall stations / water level stations.

  The data aggregating server 33 narrows down the camera stations in the vicinity of the rainfall stations / water level stations recorded in the list of dangerous rainfall stations / water level stations using the latitude / longitude information as a key. In addition, the data aggregation server 33 narrows down the camera stations that are related (influential) to the narrowed-down camera stations. The data aggregation server 33 is a narrowed-down camera station, and creates an extracted camera station DB, which will be described later, which is a list of camera stations to be processed by the image processing server 34. For example, when the upstream camera station is included in the list of dangerous rainfall stations / water level stations, the data aggregation server 33 narrows down the camera stations near the downstream hydrophilic facility. The data aggregation server 33 notifies the image processing server 34 of a list of camera stations to be processed by the image processing server 34.

  In step S14, the image processing server 34 performs camera angle-of-view specifying processing. The camera angle-of-view specifying process is a process for preset-controlling a camera station to be processed to an angle of view centered on a point where the camera operation is the most.

  The image processing server 34 sequentially selects camera stations to be processed based on the list of camera stations to be processed notified from the data aggregation server 33. The image processing server 34 requests the camera monitoring server 32 to perform preset control on the selected camera station to the angle of view centered on the point where the camera operation is the largest. Note that the preset control to the angle of view centered on the point where the camera operation is the most can be realized by using a camera station PTZ information DB 38 described later. The camera monitoring server 32 executes processing for preset control of the camera station to the angle of view requested from the image processing server 34.

  The preset control is designated using a preset ID. The camera monitoring server 32 executes processing for preset control using control information associated with the preset ID of the camera station preset information DB in FIG. The captured image is transmitted from the preset-controlled camera station.

  In step S15, the image processing server 34 performs image processing on the image received from the preset-controlled camera station. The image processing server 34 detects a person by performing image processing on an image taken at a preset position (view angle) set by the camera view angle specifying process.

  When there is no person detection record, the image processing server 34 executes image processing of a block including a point (X coordinate, Y coordinate) with the largest number of camera operations. The image processing server 34 sets the block as the first processing range, and expands the processing range radially, for example. When there is a person detection record, the image processing server 34 performs image processing from the processing range with the highest detection record. The image processing server 34 expands the processing range radially. The processes in steps S14 and S15 are repeated for the number of camera stations included in the list of camera stations to be processed.

  In step S16, the image processing server 34 performs detection data update / analysis processing. The image processing server 34 stores the image processing result of step S15 in a camera station detection point DB 40 described later. For example, the image processing server 34 accumulates detection points for each angle of view of each camera station in the camera station detection point DB 40. The image processing server 34 analyzes the processing range for each angle of view of each camera station based on the accumulated camera station detection point DB 40, and updates a later-described camera station processing range DB 39.

  In step S <b> 17, the image processing server 34 displays the image processing result on the PC 35 and the TV 37 which are examples of management content. Therefore, the river manager can confirm the image processing result displayed on the PC 35 and the TV 37.

  FIG. 7 is a configuration diagram of an example of the camera station PTZ information DB. The camera station PTZ information DB 38 has data reception date, camera ID, camera name, camera station location (latitude, longitude), preset ID, control information (X coordinate, Y coordinate, zoom value), and video multicast address as data items. The camera station PTZ information DB 38 stores the PTZ information received from the camera monitoring server 32 as control information together with the data reception date and time.

  FIG. 8 is a configuration diagram of an example of the extracted camera station DB. The extracted camera station DB has data items such as camera ID, camera name, camera station location (latitude, longitude), and video multicast address. The extracted camera station DB obtains the latitude / longitude information from the camera station DB of FIG. 4 (A), the latitude / longitude information of the rain station / water level station near or related to the rain station / water level station list. It is narrowed down as a key.

  FIG. 9 is a configuration diagram of an example of the camera station processing range DB. The camera station processing range DB 39 in FIG. 9 includes data items such as camera ID, camera name, camera station location (latitude, longitude), angle of view information (X coordinate, Y coordinate, zoom value), and processing range reference point (X coordinate, Y Coordinate), processing range (X value, Y value), and video multicast address.

  FIG. 10 is a configuration diagram of an example of the camera station detection point DB. 10 includes data reception date and time, camera station location (latitude, longitude), field angle information (X coordinate, Y coordinate, zoom value), and a plurality of detection points (X of the processing range reference point). Coordinate and Y coordinate, and X value and Y value of the processing range).

  FIG. 11 is an image diagram of an example of the angle of view specified by the angle of view information (X coordinate, Y coordinate, zoom value). The angle-of-view information (X coordinate, Y coordinate) represents a position (point) 51 that was most often selected by camera operation. The angle of view information (zoom value) represents the width 52 and the height 53 of the angle of view 54.

  FIG. 12 is an image diagram of an example of the processing range specified by the processing range reference point (X coordinate, Y coordinate) and the processing range (X value, Y value). As shown in FIG. 12, the processing range reference point (X coordinate, Y coordinate) represents the processing range reference point 61 of the processing range 64. The processing range (X value, Y value) represents the width 62 and height 63 of the processing range 64. In the camera station detection point DB 40 of FIG. 10, the detection points are represented by a processing range reference point 61 and a width 62 and a height 63 of the processing range 64 as shown in FIG.

  FIG. 13 is an image diagram of an example in which a plurality of detection points are included in one angle of view. As shown in FIG. 13, when a plurality of detection points 72 and 73 are included in one angle of view 71, a processing range reference point (X coordinate, Y coordinate) and a processing range (X value, Y value) is set.

  FIG. 14 is a flowchart of an example of the first camera angle-of-view specifying process. In step S21, the image processing server 34 performs a camera specifying process. In the camera specifying process, one camera station that executes image processing is specified from the list of camera stations to be processed in the area limiting process in step S13.

  In step S <b> 22, the image processing server 34 performs the most-view angle specifying process for the camera station specified in step S <b> 21. For the process of specifying the most angle of view, the camera station PTZ information DB 38 is referred to, and the position (PTZ information) most frequently selected in the camera station specified in step S21 is specified. That is, the most-view angle specifying process specifies the angle of view that was most frequently selected in the camera station specified in step S21.

  In step S23, the image processing server 34 performs camera preset control processing for the camera station specified in step S21. The camera preset control process requests the camera monitoring server 32 to perform preset control to the angle of view specified in step S22.

  Therefore, the camera monitoring server 32 can execute processing for preset control of the camera station to the angle of view requested from the image processing server 34. The captured image is transmitted from the preset-controlled camera station.

  The most frequent angle-of-view specifying process in step S22 is based on the following grounds. The angle of view in which the river manager operates the camera during operation and monitors more is an angle of view that is important in normal monitoring. Similarly, even at the time of water increase, the angle of view to be confirmed first is the most frequently observed angle of view. In this way, it is considered that priority is given to performing image processing from the largest angle of view that was most often selected by camera operation, because even river managers are required to respond quickly to initial movements when water increases. .

  FIG. 15 is a flowchart of an example of the camera angle-of-view specifying process after the second time. In step S24, the image processing server 34 performs a camera specifying process. In the camera specifying process, one camera station that executes image processing is specified from the list of camera stations to be processed in the area limiting process in step S13.

  In step S25, the image processing server 34 performs an angle-of-view specifying process for the camera station specified in step S24. In the second view angle specifying process, the camera station PTZ information DB 38 is referred to, and the position (PTZ information) that was most often selected second in the camera station specified in step S24 is specified. In other words, the process for specifying the angle of view after the nth time refers to the camera station PTZ information DB 38 and specifies the position (PTZ information) that is frequently selected nth in the camera station specified in step S24.

  In step S26, the image processing server 34 performs camera preset control processing on the camera station specified in step S24. The camera preset control process requests the camera monitoring server 32 to perform preset control to the angle of view specified in step S25.

  Therefore, the camera monitoring server 32 can execute processing for preset control of the camera station to the angle of view requested from the image processing server 34. The captured image is transmitted from the preset-controlled camera station.

  In the processing with only the maximum angle of view, processing is performed at one position for each camera station. However, in the camera angle-of-view specifying process that is repeatedly performed, it is possible to execute image processing over a wide area with respect to one camera station. Therefore, it is possible to detect a person in the river water facility more extensively in the camera angle-of-view specifying process that is repeatedly performed.

  FIG. 16 is a flowchart of an example of image processing. In step S31, the image processing server 34 performs background photographing processing on the image received from the preset-controlled camera station. In order to perform background difference processing, which will be described later, the background photographing processing performs processing for photographing an image received from a preset-controlled camera station as a background image.

  In step S32, the image processing server 34 performs a reference process of the camera station detection point DB 40. The reference process of the camera station detection point DB 40 executes a process of referring to the camera station detection point DB 40 for the presence / absence of a result (person detection result) of detecting a person in the past by the camera station executing the background difference process. The camera station detection point DB 40 may store human detection results separately in, for example, summer and winter.

  In step S33, the image processing server 34 performs detection result determination processing. The detection result determination process executes a process of determining the presence or absence of a detection result based on the reference result of the camera station detection point DB 40 in step S32.

  If there is a detection record, the image processing server 34 performs a range specification process (radial enlargement process) with a detection record in step S34. The image processing server 34 executes processing for designating a range in which image processing is performed in a processing range having a high detection record.

  The image processing server 34 can shorten the time until the person is detected by starting the image processing from the processing range having the detection results. When the person cannot be detected by the background difference process described later, the image processing server 34 performs a process of expanding the processing range radially, for example, in units of one block, and designating a range for image processing.

  If there is no detection record, the image processing server 34 performs a range designation process (radial enlargement process) without a detection record in step S35. The image processing server 34 refers to the camera station PTZ information DB 38. The image processing server 34 executes processing for designating a range in which image processing is performed in the smallest block including the position (X coordinate, Y coordinate) 51 that is most often selected by camera operation.

  The image processing server 34 can shorten the time until the person is detected by starting the image processing from the processing range including the position 51 that is most often selected by the camera operation. When the person cannot be detected by the background difference process described later, the image processing server 34 performs a process of expanding the processing range radially, for example, in units of one block, and designating a range for image processing.

  In step S36, the image processing server 34 performs background difference processing. The image processing server 34 compares the background image captured in step S31 with the current image captured during the background difference process, and executes a process of extracting a difference amount (change amount). The image processing server 34 detects a person from the extracted difference amount. The image processing server 34 repeats the process of step S34 or S35 until a person is detected or the processing range cannot be expanded.

  When the image processing server 34 detects a person or cannot expand the processing range, the image processing server 34 performs determination result processing in step S37. In the determination result process, for example, JPEG data that is a result of the background difference process is displayed on the PC 35 and the TV 37, which are examples of management content.

  It is also conceivable that the image processing server 34 periodically shoots and accumulates background images instead of the background shooting process in step S31. However, it is conceivable that the background image has a greater change in state such as a shadow as there is a difference in time.

  For example, even if the background is photographed every 4 hours, the length of the shadow in the background image may change after 2 hours. Therefore, since the image processing server 34 may cause a false detection due to a change in a situation such as a shadow in the background image, the image processing server 34 is configured to process the background image immediately before performing the image processing.

  FIG. 17 is a flowchart of an example of detection data update / analysis processing. In step S41, the image processing server 34 updates the camera station detection point DB 40. In the update process of the camera station detection point DB 40, the update process of the camera station detection point DB 40 is executed based on the result of the background difference process in step S36.

  In step S42, the image processing server 34 performs detection point analysis processing. The detection point analysis process refers to the camera station detection point DB 40 and executes a process for specifying a point where a person is frequently detected in each camera station.

  In step S43, the image processing server 34 updates the camera station processing range DB 39. In the update processing of the camera station processing range DB 39, the camera station processing range DB 39 is updated at points (person detection points) where there are many people detected in each camera station specified by the detection point analysis processing.

  Since the camera station processing range DB 39 has two DBs of time (summer season and winter season), it is also possible to determine summer and winter seasons using the reception date and time of stored data as a key. This is because it is assumed that the person detection point changes depending on the season even at the same angle of view of the camera station.

  FIG. 18 is an image diagram of an example of a range designation process without detection results. The image processing server 34 expands the processing range and repeats processing each time the image processing is repeated, centering on the block including the position (X coordinate, Y coordinate) 51 that is most often selected by camera operation. Early human detection becomes possible.

  Note that when there is no detection result, the image processing server 34 starts the image processing from the position 51 that is most often selected by the camera operation, so that the target moving object (person) is often present. Image processing can be performed from the center, and moving objects can be detected more quickly. In addition, it is considered that the vegetation that becomes a disturbance is often concentrated outside (upper and lower) of the angle of view. Accordingly, the image processing server 34 performs image processing from the center of the angle of view, thereby reducing the influence of disturbance caused by plants and the like that are often outside the angle of view, and reducing the false detection rate due to disturbance. Can be lowered.

  FIG. 19 is an image diagram of an example of a range designation process with detection results. The image processing server 34 can detect a person at an early stage by expanding the processing range and repeating the processing each time the image processing is repeated, centering on the processing range where the detection results have been high.

  Note that when there is a detection record, the image processing server 34 starts image processing from a processing range where the detection record is high, so that image processing can be performed from a processing range in which there are many moving objects (persons) as targets. It is possible to detect a moving object earlier. In addition, since the moving object can be detected earlier, the image processing server 34 can reduce the influence of disturbance due to plants and the like, and can reduce the false detection rate due to disturbance.

  The image processing server 34 according to the present embodiment selects which CCTV camera 21 by properly using the processing range for starting the image processing depending on whether there is no detection result or no detection result for the angle of view of the CCTV camera 21. The moving object (person) can be detected more quickly for the angle of view. In addition, since the image processing server 34 can detect a moving object earlier, the false detection rate due to disturbances such as vegetation can be reduced.

  FIG. 20 is an image diagram of an example of detection point analysis processing. FIG. 20 shows an example of an angle of view 100 to 105 including detection points. The image processing server 34 narrows down the detection points having a large number of detection results as indicated by the angle of view 105 from the detection points included in the angle of view 100 to 105.

  Note that the image processing server 34 determines the first processing range at a plurality of locations, such as when there are many detection results at a plurality of detection points, or when the detection results of the single unit are the same and the processing ranges are different. For example, when the detection results of the pattern of the angle of view 103 are increased, the first processing range is the angle of view 106.

  FIG. 21 is an image diagram of an example of a screen displayed in the management content. FIG. 21A is an image diagram of a screen displaying a list of image processing results. By selecting the CCTV camera 21 from the list on the screen of FIG. 21A and pressing the processing result display button 111, the result of moving object detection as shown in FIG. 21B is displayed on the management content. FIG. 21B shows the detected moving object (person) and the processing ranges 113 and 114.

  When the CCTV camera 21 is selected from the screen list of FIG. 21A and the video display button 112 is pressed, the management video contains the current video shot by the CCTV camera 21 as shown in FIG. Is displayed.

  The image processing server 34 is configured by a PC 80 as shown in FIG. 22, for example. FIG. 22 is a hardware configuration diagram of an example of a PC.

  22 includes an input device 81, a display device 82, and a PC main body 83. The PC main body 83 includes a main storage device 91, an arithmetic processing device 92, an interface device 93, a recording medium reading device 94, and an auxiliary storage device 95 that are connected to each other via a bus 97. An input device 81 and a display device 82 are connected to the bus 97.

  The input device 81, the display device 82, the main storage device 91, the arithmetic processing device 92, the interface device 93, the recording medium reading device 94, and the auxiliary storage device 95 connected to each other via the bus 97 are managed by the arithmetic processing device 92. Data can be sent and received between each other. The arithmetic processing unit 92 is a central processing unit that controls operation of the entire PC 80.

  The interface device 93 receives data from the IP network 4 or the like and passes the contents of the data to the arithmetic processing device 92. The interface device 93 transmits data to the IP network 4 or the like in response to an instruction from the arithmetic processing device 92.

  The auxiliary storage device 65 stores a moving object detection program that causes the PC 80 to execute at least processing in the image processing server 34 as a part of a program that causes the PC 80 to perform the same function as the image processing server 34. Then, the arithmetic processing device 92 reads out the moving object detection program from the auxiliary storage device 95 and executes it, whereby the PC 80 functions as the image processing server 34. The moving object detection program may be stored in the main storage device 91 accessible to the arithmetic processing device 92. The input device 81 receives data input under the control of the arithmetic processing device 92. The moving object detection program can be recorded in a recording medium 96 that can be read by the PC 80.

  Examples of the recording medium 96 include a magnetic recording medium, an optical disk, a magneto-optical recording medium, and a semiconductor memory. Magnetic recording media include HDDs, flexible disks (FD), magnetic tapes (MT) and the like. Examples of the optical disc include a DVD (Digital Versatile Disc), a DVD-RAM, a CD-ROM (Compact Disc-Read Only Memory), and a CD-R (Recordable) / RW (ReWriteable). Magneto-optical recording media include MO (Magneto-Optical disk).

  When distributing the moving object detection program, it is conceivable to sell a portable recording medium 96 such as a DVD or a CD-ROM in which the moving object detection program is recorded. For example, the PC 80 that executes the moving object detection program reads the moving object detection program from the recording medium 96 in which the recording medium reader 94 has recorded the moving object detection program. The arithmetic processing device 92 stores the read moving object detection program in the main storage device 91 or the auxiliary storage device 95.

  The PC 80 reads the motion detection program from the main storage device 91 or the auxiliary storage device 95 that is its own storage device, and executes processing according to the motion detection program. The arithmetic processing unit 92 implements various processes as described above according to the moving object detection program.

  The system including the moving body detection apparatus of the present embodiment is assumed to be used as described below, for example. For example, it is assumed that water increases in a river and a system including the moving object detection device of this embodiment detects a person in a hydrophilic facility. The river manager can view the image processing results on the PC 35 and check the person in the hydrophilic facility.

  Also, the TV 37 displays a current video in which a person in a hydrophilic facility is photographed by the CCTV camera 21, for example. For example, river managers can use existing broadcasting systems and warning equipment to evacuate from speakers installed along rivers and alert people by turning on warning lights to help people in dangerous hydrophilic facilities. Can evacuate.

The present invention may have the following configurations as described below.
(Appendix 1)
A moving object detection device for detecting a moving object from an image,
Means for photographing with priority on the angle of view with a high photographing frequency for each camera with reference to the first storage unit that stores the camera control results for each camera;
A moving body detection apparatus comprising: means for detecting a moving body with priority given to a range having a moving body detection result with reference to a second storage unit that stores the moving body detection result for each camera.
(Appendix 2)
The means for detecting a moving object starts moving object detection from a range having a track record of moving object detection if the captured image includes a range having a track record of moving object detection, and detects a moving object while expanding the range having the track record of moving object detection. The moving object detection device according to appendix 1, wherein:
(Appendix 3)
The moving object detection means starts moving object detection from a range at the center of the captured image if the captured image does not include a range with a track record of moving object detection, and detects the moving object while expanding the range at the center. The moving object detection device according to appendix 1 or 2, wherein:
(Appendix 4)
The moving object detection according to any one of claims 1 to 3, further comprising means for updating the moving object detection result stored in the second storage unit based on a moving object detection result by the moving object detection unit. apparatus.
(Appendix 5)
The said 2nd memory | storage part divides the said moving body detection performance according to time, and memorize | stores it for every camera, The moving body detection apparatus as described in any one of the supplementary notes 1 thru | or 4 characterized by the above-mentioned.
(Appendix 6)
To the computer that detects the moving object from the image,
With reference to the first storage unit that stores the camera control results for each camera, each camera is allowed to shoot with a higher angle of view,
A moving object detection program that executes a process of detecting a moving object with priority given to a range having a moving object detection result with reference to a second storage unit that stores the moving object detection result for each camera.

  The recommendation program in this embodiment can be provided not only by package software but also by a WEB service or the like. The means for photographing with priority given to the angle of view with high photographing frequency for each camera described in the above claims corresponds to the processing of step S14, and the means for detecting moving object with priority given to the range having a motion detection result is shown in step S15. This means corresponds to the process, and the means for updating the moving object detection result corresponds to the process in step S16.

DESCRIPTION OF SYMBOLS 1 Sensor equipment 2 CCTV camera equipment 3 River manager equipment 4 IP network 11 Rain gauge 12 Water level gauge 13 Flood sensor 14 IP node 21 CCTV camera 22 Encoder device 23 IP node 31 River data server 32 Camera monitoring server 33 Data aggregation server 34 Image Processing server 35, 80 PC
36 Decoder device 37 TV
38 Camera station PTZ information database (DB)
39 Camera station processing range DB
40 Camera station detection point DB
81 Input Device 82 Display Device 83 PC Main Body 91 Main Storage Device 92 Arithmetic Processing Device 93 Interface Device 94 Recording Medium Reading Device 95 Auxiliary Storage Device 96 Recording Medium 97 Bus

Claims (6)

A moving object detection device for detecting a moving object from an image,
An angle of view corresponding to the maximum imaging frequency indicated by each camera control result for each target camera with reference to the first storage unit that stores the camera control result for each camera according to detection of a dangerous state based on the sensing data Means for instructing each camera to take an angle of view setting instruction ,
With reference to the second storage unit that stores the moving object detection results according to the angle of view for each camera , the moving object is detected at the highest frequency indicated by the moving object detection results corresponding to the set angle of view for each target camera . Means for preferentially setting the range to be a motion detection target range ;
A moving object detection apparatus. The means for preferentially setting the range in which the moving object is detected as a moving object detection target range starts moving object detection from the range in which the moving object is detected if the captured image includes a range in which the moving object is detected, The moving object detection device according to claim 1, wherein the moving object detection is performed while expanding a range in which the moving object detection results are present. The means for preferentially setting the moving object detection range as a moving object detection target range starts moving object detection from a range at the center of the captured image if the captured image does not include a range with a history of moving object detection. The moving object detection device according to claim 1, wherein the moving object detection is performed while enlarging a range in the center. The apparatus further comprises means for updating the moving object detection result stored in the second storage unit based on a moving object detection result by the means for preferentially setting the moving object detection range to the moving object detection target range. The moving object detection device according to claim 1. 5. The moving object detection apparatus according to claim 1, wherein the second storage unit stores the moving object detection results for each camera by dividing the results. To the computer that detects the moving object from the image,
An angle of view corresponding to the maximum imaging frequency indicated by each camera control result for each target camera with reference to the first storage unit that stores the camera control result for each camera according to detection of a dangerous state based on the sensing data Instruct the camera to set the angle of view ,
With reference to the second storage unit that stores the moving object detection results according to the angle of view for each camera , the moving object is detected at the highest frequency indicated by the moving object detection results corresponding to the set angle of view for each target camera . The range is preferentially set as the motion detection target range .
A motion detection program that executes processing.