Abnormality detection system

Abstract

Provided is an abnormality detection system that can reduce the scale of the system and maintain or improve the parking environment. 【Solution means】The abnormality detection system (100) includes a camera (1) installed in the parking lot for acquiring images in the parking lot, a parking lot facility (3) installed in the parking lot and serving as a facility for controlling vehicles, and a control unit (2) that detects abnormalities in the parking lot based on the images in the parking lot and controls the parking lot facility (3) according to the detected abnormalities.

Description

Technical Field

[0001] The present disclosure relates to an abnormality detection system that operates parking lot facilities in response to an abnormality detected in a parking lot.

Background Art

[0002] For example, parking lots in shopping malls, high-rise buildings, etc. have many blind spots and are places where people and vehicles are mixed, so the risk of contact is high. In addition, since a parking lot is a place where vehicles carrying combustibles such as gasoline and lithium-ion batteries are concentrated, there is a risk of significant damage due to the occurrence of a fire.

[0003] Therefore, when an abnormality occurs in a parking lot, it is required to quickly detect the occurred abnormality.

[0004] As a conventional device for detecting an abnormality in a parking lot, the following is disclosed (for example, see Patent Document 1). When the device according to Patent Document 1 detects an abnormality using a parking lot camera, it communicates with the vehicles parked around the place where the abnormality is detected, and collects and analyzes information from the out-of-vehicle monitoring sensors of the vehicles.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] In the device disclosed by Patent Document 1, it is necessary to collect information for analysis from the parked vehicles. Therefore, when there are no vehicles in the parking lot, information cannot be collected.

[0007] In addition, the device in Patent Document 1 assumes that the vehicle is equipped with an out-of-vehicle monitoring sensor. Therefore, since it is necessary to construct a communication system for collecting information from the vehicle, it becomes a large-scale and limited system.

[0008] The present disclosure has been made to solve the above problems, and an object thereof is to provide an abnormality detection system capable of suppressing the scale of the system and maintaining or improving the usage environment of the parking lot.

Means for Solving the Problems

[0009] The abnormality detection system according to the present disclosure includes an imaging unit that images the inside of the parking lot, parking lot facilities that are installed in the parking lot and are facilities for controlling vehicles, and an abnormality inside the parking lot based on the video imaged by the imaging unit. A control unit that detects and controls the parking lot facilities according to the detected abnormality, and the control unit determines at least one of speeding violations, stop sign violations, reverse driving violations outside of parking, reverse driving on the wrong route, and unauthorized entry of general vehicles into the loading area as a rule defined in the parking lot, and as an abnormality detection, detects a vehicle that violates the rule defined in the parking lot, and stores information on whether or not the vehicle violates the rule. 、

Function

Effects of the Invention

[0010] According to the present disclosure, it is possible to provide an abnormality detection system capable of suppressing the scale of the system and maintaining or improving the usage environment of the parking lot.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Mode for Carrying Out the Invention

[0012] Hereinafter, a preferred embodiment of the abnormality detection system of the present disclosure will be described with reference to the drawings.

[0013] The abnormality detection system of the present disclosure is characterized in that it detects an abnormality based on an image captured by a camera and controls the operation of the parking lot facilities according to the content of the detected abnormality.

[0014] In addition, in the abnormality detection system of the present disclosure, by using a camera installed in the parking lot, the system can be constructed at a lower cost.

[0015] By utilizing the abnormality detection system of the present disclosure, it is possible to quickly respond to abnormalities occurring in the parking lot, which have been difficult to recognize until now, and to take post-event measures.

[0016] Here, "abnormality" comprehensively includes the occurrence of fire or smoke, contact between people and vehicles, contact between vehicles, vehicle theft, vandalism in the vehicle, violation of on-site rules, falling objects, and human falls, etc., and is not limited to specific events.

[0017] Embodiment 1. FIG. 1 is a diagram showing the overall configuration of a parking lot according to Embodiment 1 of the present disclosure, and FIG. 2 is a front view showing the vicinity of area R shown in FIG. 1.

[0018] The parking lot shown in FIG. 1 has a plurality of vehicle lanes 103 through which people or vehicles 200 pass, and a plurality of vehicle compartments 111 for parking vehicles 200. A plurality of cameras 1 are installed at various locations in the parking lot as imaging units for vehicle lane control purposes.

[0019] As shown in FIG. 1, the cameras 1 are provided at various locations in the parking lot to detect the vehicles 200 traveling on the vehicle lanes 103 and the vehicles 200 parked in the vehicle compartments 111. In addition, as shown in FIG. 2, a 360-degree camera as the camera 1 may be provided at a position where the whole can be overlooked, for example, at the ceiling portion in the center of the parking lot.

[0020] The cameras 1 use cameras that are already used for existing vehicle detection and license plate recognition, but may also be newly installed cameras.

[0021] Near the entrance and exit of the parking lot, an entrance car gate 3A, an exit car gate 3B, a ticket issuing facility 3C, and a fee settlement facility 3D are provided. Note that depending on the parking lot, the entrance car gate 3A, the ticket issuing facility 3C, and the fee settlement facility 3D may be absent, or only some of them may be installed.

[0022] In addition, vehicle guidance facilities 3E are provided at various locations inside the parking lot shown in FIG. 1. Here, the vehicle guidance facilities 3E are, for example, full-empty lights that guide the full-empty state of each floor, each area, or the entire parking lot by display or voice, display facilities that guide the route to the exit by display or voice, invitation lights that color-code and display the occupancy status of each vehicle compartment 111 in red, green, etc., a departure warning device installed at the entrance and exit of the parking lot, signboards, and the like.

[0023] In addition to these, pre-settlement facilities (not shown) or the like may be provided, or instead of the entrance car gate 3A and the exit car gate 3B, a flap device installed in each passenger compartment 111 may be provided.

[0024] FIG. 3 is a diagram showing a configuration example of the abnormality detection system according to Embodiment 1 of the present disclosure.

[0025] The abnormality detection system 100 is a system that detects an abnormality occurring in the parking lot shown in FIG. 1 and controls the facilities in the parking lot according to the detected abnormality, and includes a camera 1, a control unit 2, parking lot facilities 3, and a management terminal 4.

[0026] As described above, the camera 1 is an imaging unit installed at various locations in the parking lot.

[0027] Based on the video imaged by the camera 1, the control unit 2 performs object discrimination and abnormality determination, and selects and controls the target parking lot facilities 3.

[0028] The control unit 2 is a computer system equipped with a central processing unit, a storage device, a communication device, etc., and has functional units such as a determination unit 21 and a communication unit 22.

[0029] The determination unit 21 acquires the video imaged by the camera 1 and performs object discrimination processing using the AI constructed by pre-learning. In addition, the determination unit 21 acquires the coordinate information of the discriminated object and determines whether an abnormality has occurred based on the acquired information.

[0030] The communication unit 22 is a communication module that communicates with the parking lot facilities 3 and the management terminal 4 by wire or wirelessly.

[0031] The parking lot equipment 3 is equipment installed in the parking lot and operates according to the control signal from the control unit 2. The parking lot equipment 3 is equipment specific to the parking lot for controlling the vehicle 200, and includes an entrance car gate 3A, an exit car gate 3B, a ticket issuing facility 3C, a payment settlement facility 3D, a pre-payment facility, a flap device, and a vehicle guidance facility 3E as a display facility for guidance, etc.

[0032] The management terminal 4 is installed in a management room or the like, and is a computer system mainly used by an administrator who manages the parking lot.

[0033] Next, the outline of the operation of the abnormality detection system 100 will be described using the flowchart of FIG. 4.

[0034] In step S101 of FIG. 4, the camera 1 installed at various locations in the parking lot photographs the inside of the parking lot.

[0035] In step S102, the control unit 2 acquires the video captured by the camera 1, and as an abnormality determination process, first determines whether a target object exists in the video.

[0036] Here, if a suspicious person's face, the license plate number of a suspicious vehicle, fire, smoke, etc. exist in the video, the existence itself can be treated as abnormal. Therefore, the control unit 2 registers or learns such specific objects in advance, and when these exist in the video, it determines that an abnormality has occurred.

[0037] In addition, the control unit 2 determines whether the operation of the detected object is an abnormal operation or not from the coordinate information of the object by registering or learning operations such as squatting and falling in advance.

[0038] As the image processing in this step S102, in addition to the method using AI such as CNN (Convolutional Neural Network), the change in the video may be detected by the background difference method. Further, for the determination of abnormality, in addition to the method using AI such as LSTM (Long Short Term Memory), a rule-based method without using AI may also be used.

[0039] When it is determined in step S103 that an abnormality has occurred in the parking lot, the control unit 2 selects, in step S104, the parking lot facility 3 to be operated according to the content of the abnormality and controls it. On the other hand, when it is determined in step S103 that no abnormality has occurred in the parking lot, the control unit 2 ends the process of FIG. 4.

[0040] The process of FIG. 4 is repeatedly executed according to a specified cycle.

[0041] Subsequently, specific examples of the content of the abnormality detected according to Embodiment 1 will be presented, and how the abnormality detection system 100 detects the abnormality and controls the target parking lot facility 3 will be described.

[0042] <Detection of a vehicle violating the on-site rules> FIGS. 5 and 6 are diagrams showing operation examples of the parking lot facility 3 when a vehicle 200 violating the on-site rules is detected.

[0043] In step S201 shown in FIG. 5, the cameras 1 installed at various locations in the parking lot photograph the inside of the parking lot in the same manner as in step S101 of FIG. 4.

[0044] In step S202, the control unit 2 acquires the images photographed by the respective cameras 1 and inputs the images to the CNN as the abnormality determination process shown in FIG. 4. Thereby, the control unit 2 performs object discrimination as to whether an object, here a vehicle 200, exists in the image.

[0045] When the control unit 2 determines in step S203 that the vehicle 200 does not exist in the image, the process of FIG. 5 ends. On the other hand, when the control unit 2 determines that the vehicle 200 exists in the image, in step S204, it calculates the difference from the stored information.

[0046] When determining whether there is a speed violation, the stored information includes the coordinate information obtained during the processing one cycle before. The control unit 2 calculates the difference based on the obtained coordinate information of the vehicle 200.

[0047] Also, by setting a threshold value of the coordinate change amount in advance as the stored information, it is possible to calculate whether the change amount (difference) of the coordinate information of the vehicle 200 within the coordinates is less than the threshold value.

[0048] In step S205, the control unit 2 determines whether a violation of the on-site rules is detected based on the coordinate change amount calculated in step S204. This step S205 corresponds to step S103 in FIG. 4.

[0049] When the control unit 2 does not detect a violation of the on-site rules, the process of FIG. 5 ends. On the other hand, when the control unit 2 detects a violation of the on-site rules, in step S206, it controls the operation of the parking facility 3.

[0050] Here, the situation when a violation of the on-site rules is detected is shown in FIG. 6(A). The abnormality detection system 100 detects a vehicle 200A that has exceeded the speed or has not stopped temporarily as a vehicle 200 that violates the on-site rules here.

[0051] When a violating vehicle is detected, as shown in FIG. 5(B), the abnormality detection system 100 controls the operations of the fee settlement facility 3D, the vehicle guidance facility 3E, etc., to alert the violating vehicle 200A by display or voice.

[0052] Also, an image or the like of the violating vehicle 200A may be notified to the management terminal 4.

[0053] On the other hand, for those who have not violated the on-site rules, for example, they may be continuously encouraged not to violate the on-site rules by a pre-settlement facility (not shown). In addition, the anomaly detection system 100 may perform a preferential process of giving discounts on parking fees and privileges such as service tickets that can be used at the next use by the fee settlement facility 3D.

[0054] In addition to these, when the vehicle 200 enters the site, the anomaly detection system 100 may use the ticket issuing facility 3C to request not to violate the on-site rules and may also give a prior notice about discount processing of parking fees, preferential treatment, etc.

[0055] In addition to the above speeding and stop violation, the anomaly detection system 100 can also detect violations of driving rules such as reverse driving other than when parking and driving in the opposite direction of the driving route, and entry of general vehicles into the loading area as violations of the on-site rules. Similarly, when the anomaly detection system 100 detects these, it controls the fee settlement facility 3D or the vehicle guidance facility 3E to give a warning by display or voice.

[0056] In addition, the anomaly detection system 100 can also detect parking at locations other than the passenger compartment 111, such as the corners of the roadway 103. In this case, the anomaly detection system 100 may recognize the vehicle number of the vehicle 200 and notify the result to the administrator, or may give a guidance to prompt movement using display terminals inside and outside the parking lot.

[0057] <Detection of Suspicious Behavior> FIG. 7 and FIG. 8 are diagrams showing operation examples of the parking lot facility 3 when detecting suspicious behavior.

[0058] In step S301 shown in FIG. 7, the camera 1 installed at various locations in the parking lot photographs the inside of the parking lot in the same way as in step S101 of FIG. 4.

[0059] In step S302, the control unit 2 acquires the images captured by each camera 1 and inputs the images into the CNN as the abnormality determination process shown in FIG. 4. Thereby, the control unit 2 determines whether an object, here a human figure, exists in the image.

[0060] In step S303, if the human figure does not exist in the image, the control unit 2 ends the process of FIG. 7. On the other hand, if the human figure exists in the image, in step S304, the control unit 2 inputs the coordinate information indicating the human skeleton information into the LSTM.

[0061] In step S306, the control unit 2 determines whether a specific action has been detected based on the time-series analysis result. For example, when detecting a squatting state as shown in FIG. 8(A), if the action of squatting or the state of squatting is detected, step S306 determines that a specific action has been detected, resulting in an abnormality determination.

[0062] If the control unit 2 does not detect a specific action, it ends the process of FIG. 7. On the other hand, if the control unit 2 detects a specific action, in step S307, it controls the operation of the parking facility 3. Also, the control unit 2 may notify the management terminal 4 at this time.

[0063] By controlling the operation of the parking facility 3, as shown in FIG. 8(B), when the vehicle 200C on which the detected suspicious person F is riding is about to leave the parking lot, the exit car gate 3B can be kept closed until the confirmation by the attendant is completed.

[0064] Also, in the case of theft of the vehicle 200, theft acts such as vandalism in the vehicle, or mischievous acts, a mechanism for notifying the owner of the vehicle 200 where such an act occurred may be provided in the abnormality detection system 100. In this case, the license plate number of each vehicle 200 is associated in advance with contact information such as an email address and a telephone number.

[0065] In addition, the abnormality detection system 100 can contribute to subsequent investigations such as tracking by causing the control unit 2 to record images such as face information and skeletal information as information about a person who has committed a suspicious act.

[0066] When the vehicle 200C of the suspicious person F enters the parking lot again at a later date, the abnormality detection system 100 may notify the management terminal 4. Also, even in this case, when the vehicle 200C of the suspicious person F attempts to leave the parking lot, the exit car gate 3B may be kept closed so that the vehicle cannot leave.

[0067] <Detection of smoke and flames> FIGS. 9 and 10 are diagrams showing operation examples of the parking lot facility 3 when detecting flames or smoke.

[0068] In step S401 of FIG. 9, the camera 1 installed at various locations in the parking lot photographs the inside of the parking lot in the same manner as in step S101 of FIG. 4.

[0069] In step S402, the control unit 2 acquires the images photographed by each camera 1 and inputs the images to the CNN as the abnormality determination process shown in FIG. 4. Thereby, the control unit 2 can perform object discrimination as to whether or not flames or smoke exist in the image using the learned model that has learned flames and smoke.

[0070] In step S403, the control unit 2 acquires the result of the object discrimination process, and in step S404, determines whether or not flames or smoke have been detected. Here, if no flames or smoke are detected, the process of FIG. 9 ends.

[0071] On the other hand, as shown in FIG. 10(A), when the control unit 2 detects flames or smoke, it controls the operation of the parking lot facility 3 in step S405.

[0072] Through the operation control of the parking lot equipment 3, the abnormality detection system 100 can switch the display content of the vehicle guidance equipment 3E to the display for emergencies, and as shown in FIG. 10(B), can open the entrance car gate 3A and the exit car gate 3B so as to facilitate evacuation. In addition, the abnormality detection system 100 may turn off or display a full parking sign (vehicle guidance equipment 3E) installed near the entrance of the parking lot to guide people into the lot.

[0073] Note that in order to improve the detection accuracy, a thermal camera may be adopted as the camera 1 to measure the temperature. In addition, the abnormality detection system 100 may detect a sign of the occurrence of a fire or smoke (abnormal heat generation) based on temperature information such as that from a thermal camera. Also, a visible light camera and a thermal camera may be used in combination.

[0074] When such an abnormality is detected, there may be a possibility of arson or the like. Therefore, the abnormality detection system 100 can also track by face recognition by causing the control unit 2 to record the faces of people near the place where the abnormality occurred.

[0075] <Other Abnormalities> In the first embodiment, it is not limited to the above form. For example, when the abnormality detection system 100 detects a shopping cart left in the parking lot from the video of the camera 1, it may regard this as an abnormal situation and notify the management terminal 4.

[0076] In addition, the abnormality detection system 100 may use face recognition to identify the person who left the shopping cart, and when the person exits the parking lot, by controlling the fee settlement equipment 3D, the vehicle guidance equipment 3E, etc., give a warning by display or voice.

[0077] Also, the abnormality detection system 100 may identify the person who returned the abandoned shopping cart to a predetermined storage location, and at the exit, perform the same discount processing, preferential treatment, etc. as above through the fee settlement equipment 3D.

[0078] Also, by using face recognition, the anomaly detection system 100 may detect, as an anomaly, a carpooling act (an act of parking one's own vehicle in a parking lot and getting into another person's vehicle to go to another place).

[0079] <Operation of Other Parking Lot Facilities 3> As operations of the parking lot facilities 3 other than those described above, the following are conceivable.

[0080] In the case of a parking lot where a welcome light for notifying the in-vehicle state in units of the passenger compartments 111 is installed, the anomaly detection system 100 may notify the surroundings of an anomaly, such as by blinking the welcome light of the passenger compartment 111 where the anomaly has occurred. Further, the anomaly detection system 100 may notify the surroundings of the installation locations of fire extinguishing equipment, AEDs, etc. through the vehicle guidance facility 3E.

[0081] When the flap devices provided in each passenger compartment 111 are used as the parking lot facilities 3, the anomaly detection system 100 may cause all the vehicles 200 to exit by lowering all the flap plates upon detecting an anomaly. Alternatively, the anomaly detection system 100 may lower only the flap plates of some of the passenger compartments 111 near the location where the anomaly has occurred.

[0082] Regarding the control of the car gate, it is not limited to opening during an anomaly, and it is possible to set the opening and closing operations for the entrance and exit or separately for different anomalies.

[0083] For example, control such as opening only the entrance and closing only the exit, or an operation where the entrance is closed and the exit can be used after settlement, such as during full occupancy or closing of the parking lot, may be adopted.

[0084] Also, in accordance with the operation of the car gate, an announcement such as "Please wait a moment as the on-site staff is currently handling something." may be made from the settlement facility.

[0085] <Other Application Examples of the Anomaly Detection System 100> In addition, the anomaly detection system 100 can also ensure the evacuation route by sending signals to automatic doors for people to enter and exit, etc., or stop the operation of equipment by sending signals to equipment other than the parking facilities 3 such as EV charging facilities. Also, in a parking lot equipped with an automatic valet parking system or the like, it is possible to prevent the spread of damage in the event of an anomaly by sending signals to the self-driving vehicle and enabling the vehicle to drive automatically.

[0086] Note that the anomaly detection system 100 described above is premised on being used in a parking lot, but it may also be applied to places other than parking lots, such as a bicycle parking lot, a truck berth in a logistics warehouse, an automobile factory, a gas station, etc. For example, in a drive-through store, if a camera is used as a device to detect vehicles, and when the vehicle is congested up to the roadway starting from an outdoor order terminal or a place where orders are received, this can be detected as an anomaly and notified to the store, or the vehicle cabin for mobile orders can be guided by voice or the like.

[0087] When organizing the features of such an anomaly detection system 100, it has the following configuration and can achieve the effects.

[0088] The anomaly detection system 100 includes a camera 1 that images the inside of the parking lot and parking facilities 3 installed in the parking lot for controlling the vehicle 200. Further, the anomaly detection system 100 includes a control unit 2 that detects anomalies inside the parking lot based on the video imaged by the camera 1 and controls the parking facilities 3 according to the detected anomalies.

[0089] Thereby, it is possible to provide an anomaly detection system 100 that can suppress the scale of the system and maintain or improve the usage environment of the parking lot.

[0090] Also, the camera 1 and the control unit 2 utilize those already installed inside the parking lot. Thereby, the anomaly detection system 100 can be provided at a lower cost.

[0091] In addition, as abnormal detections, the control unit 2 can detect a vehicle 200 that violates the rules defined within the parking lot and store information on whether the vehicle 200 violates the rules. Further, as an abnormal detection, the control unit 2 can detect suspicious behavior in the parking lot and store information regarding the person who committed the suspicious behavior. Also, as an abnormal detection, the control unit 2 can detect a fire or smoke occurring in the parking lot.

[0092] In this way, by storing information regarding the violating vehicle and information on the suspicious person, it becomes possible to utilize them for verification and tracking.

[0093] As described above, by applying the abnormality detection system 100, various events that were previously difficult to recognize in the parking lot as abnormalities, such as acts violating the in-site rules, suspicious acts by suspicious persons, and the occurrence of fire or smoke, can be captured as abnormalities, and the control of the parking lot equipment can be performed.

Description of Reference Numerals

[0094] 1 Camera (imaging unit), 2 Control unit, 3 Parking lot equipment, 3A Entrance car gate, 3B Exit car gate, 3C Ticket issuing equipment, 3D Fare settlement equipment, 3E Vehicle guidance equipment, 4 Management terminal, 21 Judgment unit, 22 Communication unit, 100 Abnormality detection system, 103 Vehicle path, 111 Passenger compartment, 200, 200A, 200B, 200C Vehicles, F Suspicious person, R Area.

Claims

1. An imaging unit that captures an image of the parking lot; Parking lot equipment, which is equipment for controlling vehicles, installed in the parking lot; A control unit that detects an abnormality in the parking lot based on the image captured by the imaging unit and controls the parking lot equipment in response to the detected abnormality; and Equipped with The control unit is The rules set forth in the parking lot are at least one of the following: speeding, stop violations, backing up when not parked, driving the wrong way, and allowing general vehicles to enter a loading area; The system has a function of detecting a vehicle violating a rule established in the parking lot as the detection of the abnormality, storing information on whether or not the vehicle is violating the rule, and a function of controlling the parking lot equipment to issue a warning to the vehicle violating the rule. Anomaly detection system.

2. The control unit is Calculating a change amount between coordinate information of the vehicle in the image captured in the current cycle and coordinate information of the vehicle in the image captured in the processing of the previous cycle as a coordinate change amount; A vehicle violating the rule is detected based on a comparison between the coordinate change amount and a preset threshold value. The anomaly detection system according to claim 1 .

3. The imaging unit and the control unit are those already installed in the parking lot. The anomaly detection system according to claim 1 or 2.

4. The control unit further has a function of detecting suspicious behavior in the parking lot as the detection of the abnormality and storing information about a person who performed the suspicious behavior. The anomaly detection system according to claim 1 or 2.

5. The control unit further has a function of detecting a flame or smoke occurring in the parking lot as the detection of the abnormality. The anomaly detection system according to claim 1 or 2.

Images