JP7467434B2 - Security Systems - Google Patents

Description

An embodiment of the present invention relates to a security system.

In recent years, security measures at hard targets such as airports have been strengthened for the purpose of preventing terrorism. Conventional security measures are mainly X-ray inspection and visual baggage inspection. The problem here is that X-ray inspection and visual baggage inspection take time to inspect each person, causing people to stagnate and become crowded. Therefore, as one approach to alleviate people from stagnation due to inspection, a walk-through type dangerous goods detection device as shown in Patent Document 1 and Non-Patent Document 1 has been proposed. In particular, the dangerous goods detection device in Non-Patent Document 1 is mobile, so it is expected to be an effective means of strengthening security measures for soft targets with relatively light security.

JP 2001-521157 A

"EVOLV EDGE TM WEAPON AND BOMB DETECTION | NO LINES AUTOMATED AND CONSISTENT SCREENING", brochure, [online], [searched April 25, 2019], Internet <https://evolvtechnology.com/wp-content/uploads/EvolvEdge_Brochure_1Up-B.pdf>

However, walk-through type hazardous materials detection devices inspect people one by one through the gate, so even though the inspection time for each person is reduced, there are concerns that congestion will occur as people wait their turn. In addition, walk-through type hazardous materials detection devices make people who pass through the gate aware that they are being inspected, which creates a feeling of oppression.

The present invention has been made in consideration of these circumstances, and one of its objectives is to provide a security system that can test many people at the same time without them being aware that testing is taking place, does not make people feel oppressed, and can alleviate the congestion that accompanies testing.

A security system according to an embodiment includes a plurality of millimeter wave sensors, a sensor data processing unit, a camera, a video data processing unit, and a display device. The plurality of millimeter wave sensors are discretely arranged at intervals on the rear side of a structure that separates a monitored space through which a plurality of people can pass, and a combined range of the detection ranges of the sensors covers the entire monitored space . The sensor data processing unit detects a position of an object to be inspected that exists in the monitored space based on composite data obtained by combining output data of the plurality of millimeter wave sensors. The camera captures an image of the monitored space from a predetermined position. The video data processing unit generates an image in which a mark is applied to a position on the image captured by the camera that corresponds to the position of the object to be inspected detected by the sensor data processing unit. The display device displays the image generated by the video data processing unit.

FIG. 1 is a diagram for explaining an example of use of a security system according to an embodiment, and is a plan view of a facility in which the security system is installed. FIG. 2 is a diagram for explaining an example of use of the security system according to the embodiment, and is a vertical cross-sectional view taken along line AA in FIG. FIG. 3 is a block diagram showing an example of the configuration of a security system according to an embodiment. FIG. 4 is a block diagram illustrating an example of the configuration of the sensor data processing unit. FIG. 5 is a block diagram showing an example of the configuration of the video data processing unit. FIG. 6 is a diagram showing an example of a monitoring image. FIG. 7A is a diagram showing an example of a monitoring image. FIG. 7B is a diagram showing an example of a monitoring image. FIG. 8 is a block diagram showing another example of the configuration of the video data processing unit. FIG. 9 is a diagram showing an example of a monitoring image. FIG. 10 is a flowchart illustrating a processing procedure of the security system according to the embodiment.

Below, a specific embodiment of the security system according to the present invention will be described in detail with reference to the attached drawings. The security system of this embodiment is a system that monitors an object to be inspected using a millimeter wave sensor in a space through which multiple people can pass. The space is, for example, a three-dimensional space. Below, an example will be described in which the space is a three-dimensional space. The three-dimensional space to be monitored (hereinafter, this will be referred to as the "monitored space") is set, for example, in a facility that may be a target of terrorism, such as an airport. For example, when the security system of this embodiment is introduced in an airport, it is effective to set an area through which an unspecified number of people pass, such as an area that connects the airport lobby to a security area with a gate, as the monitored space.

The space monitored by the security system of this embodiment is not limited to airports, but may be set in facilities where many people gather, such as shopping malls, concert venues, schools, and churches. These facilities are also likely to become targets of terrorism, so the introduction of the security system of this embodiment is highly useful. When the security system of this embodiment is introduced for anti-terrorism purposes, the objects to be inspected are firearms, firearms, knives, explosives, etc.

In addition to countering terrorism, the security system of this embodiment is also effective as a countermeasure against information leaks in, for example, office buildings where business activities are carried out. That is, in office buildings where business activities are carried out, the bringing in of camera-equipped mobile phones, smartphones, etc. may be prohibited in order to prevent the leakage of confidential information. In this case, there is a risk of congestion if baggage inspections are carried out on all people who visit the office building. Therefore, it is effective to introduce the security system of this embodiment into office buildings where business activities are carried out and monitor the bringing in of camera-equipped mobile phones, smartphones, etc. by using the entrances of the office buildings as monitored spaces.

The security system of this embodiment does not aim to accurately find the object to be inspected, but to identify a person suspected of possessing the object to be inspected and to inform the person of the person to a security guard or the like. If the person suspected of possessing the object to be inspected can be informed to the security guard, the security guard can guide the person to a place where a visual body inspection, baggage inspection, X-ray inspection, etc. will be performed, and a detailed inspection can be performed to see if the person actually possesses the object to be inspected. In this case, the freedom of people other than the person suspected of possessing the object to be inspected is not restricted, and they can enter the security area of an airport, for example. In this way, the security system of this embodiment is configured to perform a simple inspection of multiple people in the monitored space, identify a person suspected of possessing the object to be inspected, and guide only that person to a detailed inspection, so that the congestion associated with the inspection that is a concern in the conventional technology can be alleviated.

The security system of this embodiment is configured to place multiple millimeter wave sensors discretely behind structures such as wall panels and ceiling panels that separate the above-mentioned monitored space, and to monitor the monitored space using output data from these multiple millimeter wave sensors. This ensures that people in the monitored space are not made aware that an inspection is being conducted, and they do not feel oppressed. Below, the security system of this embodiment will be described in detail, assuming an example in which an area connected to a security zone in a facility with a security zone, such as an airport, is set as the monitored space.

Figures 1 and 2 are diagrams illustrating an example of the use of the security system of this embodiment, with Figure 1 being a plan view of a facility in which the security system has been installed, and Figure 2 being a longitudinal cross-sectional view taken along line A-A in Figure 1. In the example shown in Figures 1 and 2, the monitored space 100 is set in an area connected to a security zone 110 of the facility. The security zone 110 is an area where a security guard 200 stays to ensure safety.

A number of millimeter wave sensors 10 are discretely arranged on the back side of structures such as wall panels 101 and ceiling panels 102 that form the boundaries of the monitored space 100. The millimeter wave sensor 10 is a sensor that irradiates radio waves in a frequency band such as 24 GHz to 100 GHz forward and acquires information such as the distance to an object, the angle, and the reflection intensity based on the results of receiving the reflected waves reflected by the object in front. Generally, millimeter waves are electromagnetic waves in a frequency band of 30 GHz to 300 GHz, but in this embodiment, the term millimeter wave sensor 10 also refers to sensors that handle microwaves (24 GHz or higher) that are close to millimeter waves. There are also passive millimeter wave sensors 10 that receive millimeter waves emitted from objects without irradiating radio waves, and these may be used, but in this embodiment, it is assumed that an active millimeter wave sensor 10 is used that irradiates radio waves and receives reflected waves.

The multiple millimeter wave sensors 10 are discretely arranged on the back side of structures such as wall panels 101 and ceiling panels 102 that form the boundaries of the monitored space 100 so that a large detection range that covers the entire monitored space 100 is generated by integrating the detection ranges of the individual sensors. Each of the individual millimeter wave sensors 10 is installed on the back side of the wall panels 101 and ceiling panels 102 so that the detection direction faces the monitored space 100, and these millimeter wave sensors 10 cannot be seen from within the monitored space 100. The dashed square in Figure 1 indicates the projection position of the millimeter wave sensor 10 installed on the back side of the ceiling panel 102 onto the floor surface.

The number of millimeter wave sensors 10 to be installed on the back side of the wall panel 101 or the ceiling panel 102 should be determined based on the relationship between the size of the monitored space 100 and the detection range of each millimeter wave sensor 10, and should be a number that can cover the entire monitored space 100 by integrating the detection ranges of each millimeter wave sensor 10.

As shown in Figure 1, between the monitored space 100 and the security area 110, there are installed a surveillance camera 20 that captures the monitored space 100, and a display device 30 such as an LCD display that displays surveillance video (described below) generated based on the video from the surveillance camera 20. The surveillance camera 20 is an example of a camera. The surveillance video is an example of a video. Figure 1 shows an example in which two surveillance cameras 20 and two display devices 30 are installed, but the number of surveillance cameras 20 and display devices 30 is arbitrary.

The surveillance video displayed on the display device 30 is viewed by a security guard 200 staying within the security area 110. When the security guard 200 views this surveillance video and finds an individual suspected of possessing an item subject to inspection, he or she guides the individual to a detailed inspection area 120 before the individual enters the security area 110. Then, in this detailed inspection area 120, a visual body inspection, baggage inspection, X-ray inspection, etc. are carried out.

Figure 3 is a block diagram showing an example of the configuration of a security system of this embodiment. As shown in Figure 3, in addition to the multiple millimeter wave sensors 10, surveillance camera 20, and display device 30 described above, the security system of this embodiment includes a sensor data processing unit 40 that processes output data from the multiple millimeter wave sensors 10, and a video data processing unit 50 that processes video data from the surveillance camera 20. Each of the multiple millimeter wave sensors 10 transmits output data to the sensor data processing unit 40 by optical wireless communication. That is, each of the multiple millimeter wave sensors 10 and the processor are connected to be able to communicate by optical wireless communication. The surveillance camera 20 transmits video data to the video data processing unit 50 by optical wireless communication. That is, the surveillance camera 20 and the processor are connected to be able to communicate by optical wireless communication.

The sensor data processing unit 40 and the video data processing unit 50 are functional units realized by a processor. For example, the sensor data processing unit 40 and the video data processing unit 50 can be realized by a general-purpose processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) executing a predetermined program. The sensor data processing unit 40 and the video data processing unit 50 may also be realized using a dedicated processor such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

The sensor data processing unit 40 detects the position of an object to be inspected that exists within the monitored space 100 based on composite data obtained by combining output data from the above-mentioned multiple millimeter wave sensors 10. In this embodiment, the sensor data processing unit 40 detects the three-dimensional position of an object to be inspected as an example. An example configuration of the sensor data processing unit 40 is shown in FIG. 4. The sensor data processing unit 40 includes a data synthesis unit 41 and a detector 42, as shown in FIG. 4, for example.

The data synthesis unit 41 synthesizes the output data of the above-mentioned multiple millimeter wave sensors 10 to generate synthetic data equivalent to sensing data for the entire monitored space 100. The output data of each of the multiple millimeter wave sensors 10 is sensing data obtained by sensing the detection range in the monitored space 100 from the installation position of each millimeter wave sensor 10. Here, since the installation positions of each of the multiple millimeter wave sensors 10 are known, the detection range in the monitored space 100 can be identified based on the installation position of the millimeter wave sensor 10, and it is known which part of the monitored space 100 is sensed by the output data of each millimeter wave sensor 10. Therefore, by synthesizing the output data of each millimeter wave sensor 10 based on the installation position of each of the multiple millimeter wave sensors 10, synthetic data equivalent to sensing data for the entire monitored space 100 can be generated.

The detector 42 receives the synthesized data generated by the data synthesis unit 41 as input, calculates the likelihood of each object existing in the monitored space 100 being an object to be inspected, and outputs the three-dimensional position of the object whose likelihood exceeds a reference value. For example, a deep neural network (DNN) trained by deep learning can be used for the detector 42. A DNN is a neural network having multiple intermediate layers between an input layer and an output layer. The DNN constituting the detector 42 is trained to output the three-dimensional position of an object existing in the monitored space 100 whose likelihood of being an object to be inspected exceeds a reference value by sequentially updating the network parameters (weights and biases in each layer) by a backpropagation method using supervised data as the learning data.

In this case, a large number of learning data are prepared, including data generated by the data synthesis unit 41 when an object to be inspected is present in the monitored space 100, to which a label indicating the three-dimensional position of the object to be inspected is added, and data generated by the data synthesis unit 41 when an object to be inspected is not present in the monitored space 100, to which a label indicating no output is added, and these are used to train the DNN that constitutes the detector 42. These learning data can be created, for example, by running the security system of this embodiment on a trial basis before actually operating it, and generating synthetic data under various conditions. Information on the three-dimensional position of an object estimated to be an object to be inspected in the monitored space 100, output by the detector 42, is input to the video data processing unit 50.

The video data processing unit 50 applies markings to positions on the video of the monitored space 100 captured by the surveillance camera 20 that correspond to the three-dimensional positions output from the detector 42 of the sensor data processing unit 40, i.e., the three-dimensional positions of the object to be inspected in the monitored space 100, to generate a surveillance video to be displayed on the display device 30. An example configuration of the video data processing unit 50 is shown in Figure 5. The video data processing unit 50 includes a coordinate conversion unit 51, a marker generation unit 52, and a surveillance video generation unit 53, as shown in Figure 5, for example.

The coordinate conversion unit 51 converts the three-dimensional position output from the detector 42 of the sensor data processing unit 40, i.e., the three-dimensional position of an object estimated to be an inspection target in the monitored space 100, into a position (two-dimensional position) in a two-dimensional coordinate system of the image of the monitored space 100 captured by the surveillance camera 20. Since the position, orientation, angle of view, etc. of the surveillance camera 20 capturing the monitored space 100 are fixed, the correspondence between the three-dimensional position in the monitored space 100 and each pixel position (two-dimensional position) on the image of the surveillance camera 20 is uniquely determined, and this correspondence can be expressed by a coordinate conversion formula. The coordinate conversion unit 51 uses this coordinate conversion formula to convert the three-dimensional position of the object estimated to be an inspection target in the monitored space 100 into a two-dimensional position of the image of the surveillance camera 20.

The marker generation unit 52 calculates the distance from the surveillance camera 20 to the three-dimensional position of the object estimated to be the inspection target in the monitored space 100 based on the three-dimensional position output from the detector 42 of the sensor data processing unit 40 and the installation position of the surveillance camera 20, which is a known value. The marker generation unit 52 then generates a marker of a size according to the calculated distance, that is, a marker whose size increases as the three-dimensional position of the object estimated to be the inspection target in the monitored space 100 approaches the installation position of the surveillance camera 20. The marker may be generated in a form that stands out when superimposed on the image of the surveillance camera 20, and the shape, color, brightness, etc. of the marker are arbitrary.

The surveillance video generating unit 53 generates a surveillance video by superimposing a marker generated by the marker generating unit 52 on the position obtained by the coordinate transformation in the coordinate transformation unit 51 on the video of the monitored space 100 captured by the surveillance camera 20. The surveillance video generated by the surveillance video generating unit 53 is displayed on the display device 30. If there is no object to be inspected in the monitored space 100 and no three-dimensional position is output from the detector 42 of the sensor data processing unit 40, the surveillance video generating unit 53 outputs the video of the monitored space 100 captured by the surveillance camera 20 as is to the display device 30. In this case, the video of the monitored space 100 captured by the surveillance camera 20 is displayed as is on the display device 30.

The sensor data processing unit 40 and the video data processing unit 50 in the security system of this embodiment repeatedly execute the above-mentioned processing at a predetermined period based on the frame period of the surveillance camera 20, such as a period that matches the frame period of the surveillance camera 20, 1/2 the frame period, 1/3 the frame period, etc. As a result, a surveillance image in which the marker moves in accordance with the movement of the image of the monitored space 100 captured by the surveillance camera 20 is generated at any time and displayed on the display device 30.

Figure 6 is a diagram showing an example of a surveillance image 60 displayed on the display device 30. When an object presumed to be an object to be inspected is present in the monitored space 100, a marker 70 is superimposed on the surveillance image 60 displayed on the display device 30 at a position corresponding to the three-dimensional position of the object presumed to be an object to be inspected, as shown in Figure 6. Therefore, a security guard 200 who refers to this surveillance image 60 can identify a person 90 who is suspected of carrying an object to be inspected from among the people in the monitored space 100, and can guide the person 90 to a detailed inspection area 120 to perform a visual physical inspection, baggage inspection, X-ray inspection, etc.

In the image of the monitored space 100 captured by the surveillance camera 20, if multiple people overlap in the depth direction of the monitored space 100 as viewed from the installation position of the surveillance camera 20, the person behind cannot be clearly seen in the image because they are blocked by the person in front. Here, if the person behind is holding an object to be inspected, when a marker 70 is superimposed on a position on the image corresponding to the three-dimensional position of the object presumed to be the object to be inspected, the marker 70 may be superimposed on the person in front, as shown in Figure 7-1.

However, in this embodiment, a marker 70 of a size according to the distance between the three-dimensional position of the object presumed to be the inspection target and the surveillance camera 20 is superimposed, so that it is possible to accurately determine whether the person on whom the marker 70 is superimposed is a person 90 suspected of carrying the inspection target from the size of the marker 70 and the size of the person in the video. Also, when a person behind who is blocked by the person in front in the surveillance video 60 of FIG. 7-1 approaches the position of the surveillance camera 20, as shown in FIG. 7-2, a surveillance video 60 is displayed with a relatively large marker 70 superimposed on that person, so that this person can be identified as a person 90 suspected of carrying the inspection target.

Note that, as an example of marking, an example of superimposing a marker 70 on the image of the monitored space 100 captured by the surveillance camera 20 has been described above, but the marking method is not limited to this. For example, a person appearing in a position on the image corresponding to the three-dimensional position of an object presumed to be the object of inspection may be emphasized. An example of the configuration of the video data processing unit 50 in this case is shown in Figure 8. The video data processing unit 50 shown in Figure 8 includes a person detection and tracking unit 54 instead of the above-mentioned marker generation unit 52.

When the three-dimensional position of an object estimated to be an object to be inspected is output from the detector 42 of the sensor data processing unit 40, the person detection and tracking unit 54 analyzes the image of the monitored space 100 captured by the surveillance camera 20 to detect a person, and identifies the person who appears at the position obtained by the coordinate transformation in the coordinate transformation unit 51 as the person to be emphasized. Here, since there may be a person who is occluded by a previous person at the position obtained by the coordinate transformation in the coordinate transformation unit 51 as described above, it is desirable to repeatedly analyze the image of the surveillance camera 20, and identify the person as the person to be emphasized when it can be determined that the same person appears at the position obtained by the coordinate transformation in the coordinate transformation unit 51.

A method for detecting a person from a video image may use, for example, a known person detection algorithm that uses image features such as facial features. Also, a known object tracking algorithm that tracks moving objects between video frames may be used to identify a person to be highlighted from the video image of the monitored space 100 captured by the surveillance camera 20, and then track that person on the video image. In this case, the processing after identifying the person to be highlighted can be simplified.

The surveillance image generating unit 53' applies highlight processing to the person identified by the person detection and tracking unit 54 on the image of the monitored space 100 captured by the surveillance camera 20, generates surveillance image 60 in which the person is emphasized, and displays it on the display device 30. An example of the surveillance image 60 in this case is shown in FIG. 9. In this surveillance image 60, a person present at a position corresponding to the three-dimensional position of an object presumed to be an inspection object is emphasized by highlight processing 80. Therefore, a security guard 200 who refers to this surveillance image 60 can identify a person 90 who is suspected of carrying an inspection object from among the people in the monitored space 100, and guide the person to the detailed inspection area 120 to perform a visual physical inspection, baggage inspection, X-ray inspection, etc.

The highlight processing 80 may be any processing that can highlight a person 90 suspected of possessing an object to be inspected on the surveillance video 60, and may be any method, such as increasing the brightness of the area showing the person 90 to make them stand out, gradually increasing and decreasing the brightness of the area showing the person 90 to create a blinking effect, or coloring the area showing the person 90 in a specified color.

Next, the flow of operation of the security system of this embodiment will be described with reference to Figure 10. Figure 10 is a flowchart showing the processing procedure repeatedly executed at a predetermined interval by the sensor data processing unit 40 and the video data processing unit 50 in the security system of this embodiment.

When the process starts, first, the data synthesis unit 41 of the sensor data processing unit 40 synthesizes the output data of multiple millimeter wave sensors 10 that are discretely placed on the back side of the wall panels 101 and ceiling panels 102 that form the boundaries of the monitored space 100, to generate synthetic data (step S101).

Next, the composite data generated in step S101 is input to a detector 42 configured by a DNN or the like (step S102). The detector 42 receives the composite data as input, calculates a likelihood indicating the likelihood that each object present in the monitored space 100 is an object to be inspected, and is trained to output the three-dimensional position of an object whose likelihood exceeds a reference value.

Next, it is confirmed whether or not the three-dimensional position of the object presumed to be the object to be inspected is output from the detector 42 to which the composite data was input in step S102 (step S103). If the three-dimensional position of the object presumed to be the object to be inspected is not output from the detector 42 (step S103: No), the surveillance image generating unit 53 (53') of the video data processing unit 50 displays the image of the monitored space 100 captured by the surveillance camera 20 as is on the display device 30 (step S104).

On the other hand, if the detector 42 outputs the three-dimensional position of the object presumed to be the inspection object (step S103: Yes), the coordinate conversion unit 51 of the video data processing unit 50 converts the three-dimensional position output from the detector 42 into a position (two-dimensional position) in the two-dimensional coordinate system of the video of the surveillance camera 20 (step S105). Then, the surveillance video generating unit 53 (53') generates a surveillance video 60 in which a mark is applied to the position obtained by the coordinate conversion of the coordinate conversion unit 51 on the video of the surveillance camera 20, that is, the position where the object presumed to be the inspection object exists, and displays the surveillance video 60 on the display device 30 (step S106). This allows the security guard 200 to refer to the surveillance video 60 and identify a person 90 suspected of carrying an inspection object from among the people in the monitored space 100, and guide the person 90 to the detailed inspection area 120 to perform a visual physical inspection, baggage inspection, X-ray inspection, etc.

As described above in detail with specific examples, in the security system of this embodiment, a space through which multiple people can pass is set as the monitored space 100, and multiple millimeter wave sensors 10 are arranged at intervals behind structures such as wall panels 101 and ceiling panels 102 that divide the monitored space 100. Then, the output data of these multiple millimeter wave sensors 10 are synthesized to generate synthetic data, and the position of the object to be inspected present in the monitored space 100 is detected based on this synthetic data. In addition, when the monitored space 100 is photographed by the surveillance camera 20 and the position of the object to be inspected present in the monitored space 100 is detected, a predetermined marking is applied to the image of the monitored space 100 photographed by the surveillance camera 20 at a position on the image corresponding to the position of the object to be inspected, thereby generating a surveillance image 60. Then, by displaying this surveillance image 60 on the display device 30, the security guard 200 or the like who refers to the surveillance image 60 can identify a person suspected of possessing the object to be inspected.

In this way, the security system of this embodiment does not inspect each person in turn as in conventional walk-through type hazardous material detection devices, but uses multiple millimeter wave sensors 10 arranged at intervals on the back side of a structure that divides the monitored space 100 through which multiple people can pass, to identify a person suspected of possessing an object to be inspected from among multiple people present in the monitored space 100. Therefore, the security system of this embodiment can inspect many people at the same time without making them aware that an inspection is taking place, does not make people feel oppressed, and can alleviate congestion associated with inspections.

Specific embodiments of the present invention have been described above, but the above-mentioned embodiments show one application example of the present invention. The present invention is not limited to the above-mentioned embodiments as they are, and can be embodied in various ways in the implementation stage without departing from the gist of the invention. These embodiments and their variations are included in the scope and gist of the invention, and are included in the scope of the invention and its equivalents as set forth in the claims.

REFERENCE SIGNS LIST 10 millimeter wave sensor 20 camera 30 display device 40 sensor data processing unit 41 data synthesis unit 42 detector 50 video data processing unit 51 coordinate conversion unit 52 marker generation unit 53 (53') surveillance video generation unit 54 person detection and tracking unit 60 surveillance video 70 marker 80 highlight processing

Claims (6)

a plurality of millimeter wave sensors that are discretely arranged at intervals on a rear side of a structure that separates a monitored space through which a plurality of people can pass , the combined detection ranges of the respective millimeter wave sensors covering the entire monitored space;
a sensor data processing unit that detects a position of an inspection target present in the monitored space based on composite data obtained by combining output data of the plurality of millimeter wave sensors;
A camera that captures an image of the monitored space from a predetermined position;
an image data processing unit that generates an image in which a mark is applied to a position on the image captured by the camera, the position corresponding to the position of the inspection object detected by the sensor data processing unit;
a display device that displays the image generated by the image data processing unit. The security system according to claim 1, characterized in that the video data processing unit generates the video by superimposing a marker, the size of which increases as the position of the object to be inspected detected by the sensor data processing unit approaches the installation position of the camera, on the video captured by the camera. The security system according to claim 1, characterized in that the video data processing unit identifies a person appearing in a position on the video captured by the camera that corresponds to the position of the object to be inspected detected by the sensor data processing unit, and generates the video in which the person is emphasized. The security system described in claim 1, characterized in that the sensor data processing unit receives the composite data as input, calculates a likelihood indicating the likelihood that each object present in the monitored space is the object to be inspected, and detects the position of the object to be inspected present in the monitored space using a detector trained to output the position of an object whose likelihood exceeds a reference value. the sensor data processing unit receives the output data of each of the millimeter wave sensors via optical wireless communication;
The video data processing unit receives video data of the video captured by the camera via optical wireless communication.
2. The security system of claim 1. The sensor data processing unit and the video data processing unit
repeating a process including detection of the position of the inspection object and generation of the image at a predetermined period based on a period that is equal to a frame period of the camera, 1/2 the frame period, or 1/3 the frame period;
2. The security system according to claim 1.

Images