Method for predicting traffic accidents, computer device, and readable recording medium

Abstract

A method for predicting traffic accidents, a computer device, and a readable recording medium. The computer device analyzes a large number of traffic scene videos to classify the traffic scene videos in which accidents occur, and finds possible factors that lead to the accidents from each type of accident video. Then, based on the possible factors of each type of accident, all factors of the accident environment in each accident video are integrated, and various factor combinations in the same type of accident video are summarized. Based on the results of the summarized various factor combinations, a report is generated, which presents the various factor combinations in the various accident videos occurring at different road sections or intersections.

Description

Technical Field

[0001] The present invention relates to a method for identifying accident factors from traffic scene videos, and more particularly to a method for predicting traffic accidents, specifically a method for automatically sorting and summarizing accident factors based on traffic scene videos using artificial intelligence. Background Art

[0002] Nowadays, cars are often equipped with dashcams. Because drivers may not pay attention to other road details while driving, dashcams play a supporting role in recording all road conditions while driving. In addition, surveillance cameras installed at various intersections or road sections also capture traffic videos. In the event of a traffic accident, the dashcam footage and / or traffic videos captured by surveillance cameras at various intersections or road sections can help determine responsibility for the accident.

[0003] However, when people need to review a large number of traffic scene videos and identify possible factors related to each accident to clarify responsibility, the burden on relevant personnel to review the videos is undoubtedly greatly increased. It is also easy to miss or ignore some minor factors or details that may be related to the accident due to the need to quickly browse the videos. As a result, the causes of the accident may be simplified and some causes may be hidden, making it impossible to further use the true factors or details of these accidents for future prediction of intersection accidents. Summary of the Invention

[0004] Therefore, the purpose of the present invention is to provide a method for predicting traffic accidents, specifically a method for automatically organizing accident factors and predicting road accidents based on traffic scene videos, as well as a computer device and a computer-readable recording medium for implementing this method. The computer device can automatically and quickly review a large number of traffic scene videos through artificial intelligence to find all possible factors for the occurrence of accidents, and apply a combination of these factors to predict possible impending intersection traffic accidents and provide early warnings, actively eliminating some factors and reducing the risk of traffic accidents.

[0005] Therefore, the present invention provides a method for predicting traffic accidents, comprising the following steps.

[0006] (A) When an accident classification module of a computer device pre-stored with a large number of traffic scene videos determines that a collision has occurred in the traffic scene videos and confirms the existence of an accident, the traffic scene videos in which the accident occurred are classified into at least three types of accident videos: vehicle-to-person, vehicle-to-vehicle, and self-collision based on the objects of the collision and stored in a storage unit of the computer device; and the accident classification module obtains first factor data from the various accident videos within a fixed time period from a time point when the accident occurred to a fixed time point before the time point and records the data in the storage unit; the first factor data includes vehicles, pedestrians, and traffic signal equipment (traffic lights).

[0007] (B) An accident factor discovery module of the computer device reads the various accident videos from the storage unit and obtains a second factor data other than the first factor data within a fixed time period from the time point when the accident occurred to before the time point from the various accident videos; the second factor data includes geographic information and weather information.

[0008] (C) An accident summarization module of the computer device integrates all factors of the accident environment in each accident video based on the first factor data and the second factor data obtained by the accident classification module and the accident factor discovery module, and further summarizes various factor combinations in the same accident video, and generates a report based on the above-mentioned summarized various factor combinations, which presents the various factor combinations of the various accident videos occurring at different road sections or intersections.

[0009] In some embodiments of the present invention, in step (A), the accident classification module also determines that when a vehicle in one of the traffic scene videos suddenly brakes or turns, it determines that a similar accident has occurred in the traffic scene video, and classifies the traffic scene video in which the similar accident has occurred as a similar accident video; and in step (B), the accident factor discovery module also summarizes the similar accident video, and finds the second factor data in the accident environment in addition to the first factor data from the similar accident video; and the various accidents in step (C) also include similar accidents.

[0010] In some embodiments of the present invention, the method also includes an accident prediction module of the computer device detecting multiple factors appearing in an input traffic scene video to be predicted, and comparing the factors in the traffic scene video to be predicted with the factor combinations corresponding to the various accidents summarized in the report to generate a similarity value, and when it is determined that the similarity value meets a similarity threshold, generating and outputting an accident warning notification.

[0011] In some embodiments of the present invention, the method further includes the following steps.

[0012] (D) A road monitoring system receives a video of an intersection captured by a monitor installed at an intersection and provides the video to the computer device. The accident prediction module of the computer device detects multiple factors appearing in the intersection video and compares the factors in the intersection video with various combinations of factors corresponding to various accidents to generate a similarity value. When the accident prediction module confirms that the similarity value associated with the intersection video is less than the similarity threshold but greater than a first default value, a continuous intersection warning module is activated. The continuous intersection warning module determines whether a portion in which a combination of factors in the intersection video and a combination of factors corresponding to various accidents overlap includes a high-risk driving factor associated with a vehicle passing through the intersection. If so, the continuous intersection warning module generates a warning message and provides the warning message to the road monitoring system. The warning message includes the direction of travel of the vehicle.

[0013] (E) After receiving the warning message, the road monitoring system transmits a warning instruction to a warning device installed at the next intersection that the vehicle is about to arrive at based on the vehicle's travel direction contained in the warning message, so that the warning device outputs a warning message based on the warning instruction, and transmits a monitoring instruction to a monitor installed at the next intersection, so that the monitor at the next intersection takes a video of the intersection and provides it to the computer device.

[0014] (F) When the continuous intersection warning module of the computer device determines that the similarity value associated with the intersection video of step (E) still remains less than the similarity threshold but greater than the first default value, and a factor combination of the intersection video of step (E) and the repeated part of each factor combination corresponding to various accidents still include the high-risk driving factor related to the vehicle, the continuous intersection warning module generates a warning message and provides it to the road monitoring system, and the warning message includes the direction of travel of the vehicle.

[0015] (G) Repeat steps (E) and (F) until the continuous intersection warning module determines that the similarity value associated with the intersection video in step (E) is less than the first default value, or the repeated part of the factor combination of the intersection video in step (E) and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle.

[0016] In some embodiments of the present invention, in step (G), when the similarity value related to the intersection video analyzed by the continuous intersection warning module in step (E) is less than the first default value, or the repeated part of the factor combination of the intersection video in step (E) and the factor combinations corresponding to various accidents does not include the high-risk driving factor related to the vehicle, the continuous intersection warning module generates a monitoring message and provides it to the road monitoring system, and the monitoring message includes the direction of travel of the vehicle; and the method also includes the following steps after step (G).

[0017] (H) The road monitoring system transmits a monitoring instruction to a monitor installed at the next intersection that the vehicle is about to arrive at based on the vehicle's travel direction contained in the monitoring message of step (G), so that the monitor at the next intersection that the vehicle is about to arrive at takes a video of the intersection and provides it to the computer device.

[0018] (I) Repeat steps (G) and (H) until the continuous intersection warning module determines that the similarity value associated with N consecutive (N≥2) intersection videos is less than the first default value, or the repeated portion of the factor combination of the N consecutive intersection videos and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle.

[0019] In some embodiments of the present invention, the accident warning notification is also transmitted via Internet of Vehicles-related equipment to an on-board device of a vehicle that is about to pass through the traffic section displayed in the traffic scene video to be predicted.

[0020] In addition, the present invention implements a computer device for implementing the above method, including a storage unit in which a large number of traffic scene videos are pre-stored and a processing unit, the processing unit is electrically connected to the storage unit to access the storage unit, and the processing unit includes an accident classification module, an accident factor discovery module, an accident summary module, an accident prediction module and a continuous intersection warning module, and the processing unit executes the above modules to complete the method of automatically sorting out accident factors and predicting road accidents based on traffic scene videos as described above.

[0021] Furthermore, the present invention implements a computer-readable recording medium for the above-mentioned method, which stores a software program including an accident classification module, an accident factor discovery module, an accident summary module, an accident prediction module and a continuous intersection warning module. When the software program is loaded and executed by a computer device that pre-stores a large number of traffic scene videos, the computer device can complete the above-mentioned method of automatically sorting out accident factors and predicting road accidents based on traffic scene videos.

[0022] Compared to the prior art, the present invention provides a method, computer device, and readable recording medium for predicting traffic accidents. The computer device uses the accident classification module to determine whether an accident has occurred in a traffic scene video and classifies the traffic scene video in which an accident has occurred. Furthermore, the multiple accident factor discovery modules identify and statistically analyze all possible factors that may have led to the accident in accident videos of different categories. The accident summary module then summarizes the correlations between all current factor combinations that appeared in the video within a fixed time period before the accident, based on the statistically analyzed possible factors of various accidents. This allows the method to subsequently determine whether the probability of an accident occurring on a road section in a newly input traffic scene video is high, based on the dynamically changing combinations of various factors appearing in the video. This allows for early warning to eliminate or disrupt the factor combinations that may have led to the accident. Furthermore, the method automatically and rapidly examines a large number of traffic scene videos, reducing the burden on personnel searching for accident-related factors from a vast amount of video and making it easier to locate minor but potentially accident-related factors or details in the video. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Other features and effects of the present invention will be more clearly seen in the following embodiments with reference to the accompanying drawings, in which:

[0024] Figure 1 The main process steps of one embodiment of the method, computer device, and readable recording medium for predicting traffic accidents of the present invention;

[0025] Figure 2 is realized Figure 1 A schematic diagram of the software and hardware blocks mainly included in the computer device of the method flow;

[0026] Figure 3 is a schematic diagram of the computer device of this embodiment used in conjunction with the road monitoring system for road intersection monitoring and warning; and

[0027] Figure 4 and Figure 5 This is the process step of the computer device of this embodiment cooperating with the road monitoring system to monitor and warn road intersections. DETAILED DESCRIPTION

[0028] Before the present invention is described in detail, it should be noted that similar elements are denoted by the same reference numerals in the following description.

[0029] See Figure 1 FIG. 1 is a diagram showing the main steps of an embodiment of a method for predicting traffic accidents, a computer device, and a readable recording medium according to the present invention. Figure 2The computer device 1 shown is pre-stored with a large number of traffic scene videos, and these traffic scene videos can be driving recorders installed in many vehicles and / or traffic videos taken by surveillance cameras at various road sections (including intersections). These vehicles may belong to one or more public transportation fleets or transportation companies, such as but not limited to bus operating companies, taxi fleets (car dealerships), tour bus companies, gravel truck (or truck) transportation companies, etc.

[0030] The computer device 1 includes a storage unit 11, such as a memory module, and a processing unit 12, such as a central processing unit, electrically connected to the storage unit 11 to access the storage unit 11. The processing unit 12 includes an accident classification module 121, an accident factor discovery module 122, and an accident summary module 123. The modules 121 to 123 can be software programs pre-stored in the storage unit 11 and can be loaded and executed by the processing unit 12 to complete the Figure 1 The method flow shown.

[0031] Alternatively, the modules 121 to 123 may also be integrated into one or more application-specific integrated circuits (ASICs) or a programmable logic device (PLD) of the computer device 1, and the processing unit 12 is the application-specific integrated circuit chip(s) or the programmable logic device (PLD) and can complete the Figure 1 Alternatively, the modules 121-125 may be firmware recorded in a microprocessor of the computer device 1, and the processing unit 122 is the microprocessor and can execute the firmware to complete the process. Figure 1 The method flow shown.

[0032] By this, Figure 1In step S1, the accident classification module 121 reads the traffic scene videos from the storage unit 11 and identifies each traffic scene video through image recognition to determine whether there is a traffic accident in the traffic scene video, and identifies the accident as a vehicle-to-pedestrian accident, vehicle-to-vehicle accident, or self-collision accident. The traffic scene video in which the accident occurred is classified as a vehicle-to-pedestrian accident video, a vehicle-to-vehicle accident video, or a self-collision accident video and stored in the storage unit 11 according to the classification. Specifically, the accident classification module 121 uses a trained artificial intelligence algorithm to detect objects in the traffic scene video. The artificial intelligence algorithm can be, for example, but not limited to, the YOLOv4 object detection model, or other object detection models, such as, but not limited to, YOLOv1, YOLOv2, YOLOv3, CNN, R-CNN, Fast R-CNN, Faster R-CNN, and other artificial intelligence models with deep learning.

[0033] Therefore, the accident classification module 121 can detect vehicles and pedestrians (animals) in the traffic scene video and their movement trajectories, average or single vehicle speeds, surrounding fixed objects (traffic lights, dividing islands, road trees, etc.), road markings, abnormal smoke, etc. Next, the accident classification module 121 determines whether the vehicle in the video has collided, for example but not limited to detecting that the vehicle has decelerated significantly, the driving trajectory has deviated from the lane and has come to a standstill by contact with other objects, etc., and uses different identification conditions sufficient to reflect a collision to determine whether there is an accident in the traffic scene video. If there is an accident, and the accident classification module 121 further confirms that the vehicle collision object is a person (or animal), the accident classification module 121 will classify the traffic scene video from the time of the accident to the time before that. All objects detected within a fixed time period (e.g., 30 seconds) are used as first factor data, and the traffic video is classified as a vehicle-to-pedestrian accident video based on the collision relationship results, and stored together with the first factor data in the corresponding folder of the storage unit 11; similarly, the accident classification module 121 classifies traffic scene videos with accidents into vehicle-to-vehicle accident videos or self-collision accident videos in the same manner as described above, and stores these videos and the corresponding first factor data in the storage unit 11 according to the classification. It is worth mentioning that the first factor data can be used to infer data such as road conditions (e.g., whether there is a traffic jam, the current traffic flow status, vehicle or pedestrian violations), etc., and changes in vehicle type distribution (e.g., large vehicles, trucks, buses, articulated vehicles, motorcycles, etc.).

[0034] Then, if Figure 1In step S2, the processing unit 12 executes the accident factor discovery module 122. The accident factor discovery module 122 reads the various accident videos from the storage unit 11 and uses a trained artificial intelligence algorithm to search the various accident videos for second factor data (other factors) in addition to the first factor data within a fixed time period from the time of the accident to the time before the accident, such as geographic information (such as but not limited to location, road type, road attributes, etc.), traffic sign content (such as no entry, speed limit, etc.), weather information (such as sunny, cloudy, overcast, drizzle, heavy rain, fog, etc.), special conditions (such as lane reduction, diversion and closure, etc.). The artificial intelligence algorithm is, for example, but not limited to, the YOLOv4 object detection model, and may also be other object detection models, such as, but not limited to, YOLOv1, YOLOv2, YOLOv3, CNN, R-CNN, Fast R-CNN, Faster R-CNN, and other artificial intelligence models with deep learning.

[0035] Thus, as in step S2, the accident factor discovery module 122 reads the traffic scene videos (hereinafter referred to as accident videos) and the first factor data that have been classified and have occurred in the accident from the storage unit 11, identifies the accident videos, and at the same time, combines the first factor data of the accident classification module 121 to sort out the factor combination states of the first factor data and the second factor data within the fixed time before the accident occurs from the accident videos; specifically, taking the vehicle-to-pedestrian accident video as an example, the accident factor discovery module 122 will identify the second factor data in each vehicle-to-pedestrian accident video from 30 seconds before the vehicle collision to the moment of the collision, and then integrate the known first factor data to compile it into a factor combination corresponding to the vehicle-to-pedestrian accident video and store it in the storage unit 11.

[0036] Then, the accident factor mining module 122 collects the combination of factors in the collision scene from the aforementioned various accident videos and stores the result in the storage unit 11 .

[0037] Finally, if Figure 1 In step S3, the accident summarization module 123 summarizes the data of multiple factor combinations of the same type of accident videos (e.g., vehicle-to-pedestrian) stored in the storage unit 11 based on a specific factor, and unites the factor combinations that share the specific factor. For example, the factor combinations of 100 vehicle-to-pedestrian accident videos are summarized based on the road type in factor 1, and the corresponding factor combinations are united to form a summarized factor combination: a fork in the road, an urban road, 7 to 9 p.m., medium traffic, a relatively fast average speed, heavy rain, a large number of pedestrians, and pedestrians illegally crossing the intersection. When all factors in the above factor combination are fully met, it means that the probability of a collision risk in the traffic scene is the highest.

[0038] In addition, in the above-mentioned step S1, the accident classification module 121 of the present embodiment can further identify whether the vehicle in the traffic scene video has an emergency situation such as sudden braking, sudden turning or vehicle skidding when identifying each of the traffic scene videos. If so, it is determined that a similar accident has occurred in the traffic scene video, and the traffic scene video in which the similar accident has occurred is further classified as a similar accident video and stored in the storage unit 11; and as mentioned above, the accident factor discovery module 122 identifies the similar accident videos to find the second factor in the similar accident video from the similar accident videos. The data is collected, and the factor combinations of the first factor data and the second factor data in a period of time before the sudden braking, sudden turning, etc. of the vehicle in the accident-like videos are counted and stored in the storage unit 11. Similarly, the accident summarization module 123 summarizes with a specific factor, and combines the factor combinations with the same specific factor. For example, it is summarized from multiple accident-like videos to find out that if the combination of factors such as dusk, drizzle, high average speed, low traffic volume, acceleration before entering the curve or improper deceleration is met on a downhill curve, the vehicle is likely to brake suddenly or slip.

[0039] Furthermore, the accident summary module 123 can generate a report based on the above-mentioned summary results, which presents various combinations of factors in various accident environments that occurred at different road sections (intersections), and various accidents include the above-mentioned vehicle-to-person, vehicle-to-vehicle, self-collision and similar accidents. In this way, the above-mentioned public transportation fleet or transportation company can refer to the report to educate and advocate its internal vehicle drivers, or provide the report to a back-end traffic warning system for application. For example, when a vehicle travels to a certain road section that meets a certain combination of factors presented in the above-mentioned report, the traffic warning system can send an early warning notification to the vehicle-mounted device through the Internet of Vehicles related equipment or wireless network communication method to remind the vehicle driver to pay attention to the possible factors that are prone to accidents.

[0040] In addition, the processing unit 12 of this embodiment may also include an accident prediction module 124, which can detect (recognize) various factors appearing in an input traffic scene video to be predicted (e.g., a traffic video of a certain road section (intersection)) through the above-mentioned object detection model, such as geographic information (such as but not limited to location, road type), road conditions (such as many cars, few cars, traffic jam, etc.), changes in the proportion of vehicle types (such as the proportion of large trucks, trucks, buses, articulated vehicles, motorcycles, etc. in the lane), objects or signs (such as traffic regulations, traffic signs, the number and density of vehicles and pedestrians, animals, etc.), weather (such as sunny, cloudy, overcast, drizzle, heavy rain, fog, etc.), etc. ), then the accident prediction module 124 compares the factor combination in the current traffic scene with the factor combinations corresponding to the various accidents (vehicle-to-person, vehicle-to-vehicle, self-collision and similar accidents) summarized in the report of the above-mentioned accident summary module 123 to generate a similarity value, and when the similarity value between the factor combination of the traffic scene video to be predicted and the factor combination corresponding to the above-mentioned various types of accidents meets a similarity threshold (for example, these factor combinations have a high degree of overlap (for example, repetition of more than 90%) or complete overlap (complete repetition, i.e., 100%) with the factor combinations in the vehicle-to-person accident video), the accident prediction module 124 generates and outputs an accident warning notification.

[0041] In addition, if the accident prediction module 124 finds a plurality of identical highest similarity values, it may output the accident warning notice corresponding to each accident type.

[0042] The aforementioned accident warning notifications (vehicle-to-pedestrian, vehicle-to-vehicle, self-collision, and similar accidents) can be output to a traffic control center of a relevant government agency, so that relevant personnel can be mobilized to the scene of the possible impending traffic accident to direct and guide traffic flow, control road users' illegal behavior, etc., so as to eliminate some of the factors of the traffic accident, thereby reducing the similarity value of the overall factor combination. Alternatively, these accident warning notifications can be displayed on roadside or roadside electronic signs, such as electronic billboards, to remind vehicle drivers to drive carefully; or, these accident warning notifications can be transmitted via vehicle-to-vehicle network-related equipment or wireless network communication methods to on-board devices of vehicles that are about to pass through the traffic section displayed in the traffic scene video to be predicted, so as to remind vehicle drivers to pay attention to the factors that may cause accidents.

[0043] It is worth mentioning that the traffic control center can assign importance weights based on different types of traffic accidents, such as vehicle-to-pedestrian, vehicle-to-vehicle, self-collision or similar accidents. That is, when there are high accident risks at multiple intersections at the same time, the order of priority warning notifications or measures to eliminate the factors can be further determined based on the importance of the types. For example, the intersections or sections where personnel are given priority for on-site control and guidance can be determined based on the weight distribution.

[0044] In addition, if Figure 3 As shown, the computer device 1 of this embodiment can also be connected to a road monitoring system 2, wherein the computer device 1 is integrated into a monitor 31. Alternatively, the computer device 1 can be directly integrated into the road monitoring system 2 and connected to multiple external monitors. The computer device 1 can selectively monitor a specific vehicle 4 passing through continuous intersections. If the similarity value generated by the current combination of factors at the intersections does not meet the similarity threshold, but an abnormal driving factor is detected in the combination of factors at the intersections, a continuous intersection monitoring function is triggered to alert vehicles and / or pedestrians about to pass through the intersections to possible safety concerns caused by the vehicle 4. The specific implementation process is as follows: Figure 4 and Figure 5 As shown, first, Figure 4 In step S41, the road monitoring system 2 receives a signal set at an intersection, such as Figure 3 The monitor 31 of the intersection A shown in FIG. 1 captures a video of the intersection A and provides the video of the intersection A to the computer device 1. Then, as shown in FIG. Figure 4 In step S42, when the computer device 1 determines through the accident prediction module 124 that the similarity value generated by comparing a factor combination at the current intersection A with the factor combinations corresponding to the various accidents (vehicle-to-pedestrian, vehicle-to-vehicle, self-collision, and similar accidents) summarized in the report is less than the similarity threshold but greater than a first default value, the computer device 1 activates a continuous intersection warning module 125, wherein the first default value is, for example, but not limited to, 20%. For example, if the possible factors corresponding to a vehicle-to-pedestrian accident include daytime, rain, a large number of pedestrians, an intersection, and heavy traffic, and the factor combinations of the current intersection A video include nighttime, rain, a large number of pedestrians, an intersection, and the vehicle 4 speeding, then the similarity value generated by comparing the factors appearing in the intersection A video with the factor combinations corresponding to the vehicle-to-pedestrian accident is 60%, which is less than the similarity threshold but greater than the first default value.

[0045] Therefore, when the judgment result of step S42 is yes, Figure 4 As shown in step S43, the continuous intersection warning module 125 then determines whether the overlapping parts of the factor combination of the intersection A and the factor combinations corresponding to various accidents (raining, many pedestrians, intersection and the vehicle 4 speeding) include a high-risk driving factor related to the vehicle 4. In this embodiment, the high-risk driving factor is defined as but not limited to severe speeding (for example, exceeding the speed limit of the road section by more than 40 kilometers), and when the result of step S43 is yes, as shown in FIG. Figure 4 In step S44, the continuous intersection warning module 125 generates a warning message and transmits the warning message to the road monitoring system 2. The warning message includes the direction of travel, license plate information, vehicle color and model of the vehicle 4.

[0046] Then, if Figure 4 In step S45, after receiving the warning message, the road monitoring system 2 sends a warning instruction to the next intersection (e.g., the next intersection) that the vehicle 4 is about to arrive at or may arrive at according to the traveling direction of the vehicle 4 included in the warning message. Figure 3 A warning device 32 is installed at the intersection B shown in the figure, and the warning device 32 outputs a warning message according to the warning command. The warning device 32 is, for example, an electronic signboard and / or a voice output device installed at intersection B. The warning message is displayed on the electronic signboard to remind drivers of vehicles about to pass through the intersection to slow down and pay attention to oncoming vehicles, or the warning message is output through the voice output device to remind pedestrians to pay attention to vehicles. At the same time, the road monitoring system 2 transmits a monitoring command to a monitor 33 installed at the next intersection (i.e., intersection B), causing the monitor 33 at the next intersection (intersection B) to capture a video of the current intersection B and provide it to the computer device 1. The next intersection can also be multiple possible intersections predicted by the road monitoring system 2 through the image, and may not only be intersection B.

[0047] Then, if Figure 5 In step S46, the continuous intersection warning module 125 of the computer device 1 further determines whether the similarity value generated by comparing the factor combination of the intersection B with the factor combinations corresponding to various accidents continues to be lower than the similarity threshold but greater than the first default value. If the result of step S46 is yes, for example, the factor combination of the intersection B is reduced to pedestrians, crossroads, and the vehicle 4 is speeding, then the similarity value generated by comparing the factor combination of the intersection B with the factor combination corresponding to the vehicle-to-pedestrian accident is reduced to 40%, which is lower than the similarity threshold but still greater than the first default value, then the process is performed. Figure 5 In step S47, the continuous intersection warning module 125 determines whether the repeated factors (many pedestrians, intersections and the vehicle 4 overspeeding) include the high-risk driving factor generated by the vehicle, that is, the above-mentioned overspeeding. If so, Figure 5 In step S48, the continuous intersection warning module 125 generates a warning message and provides it to the road monitoring system 2. The warning message includes the traveling direction of the vehicle 4, that is, the continuous intersection monitoring is performed on the vehicle 4.

[0048] Then, the road monitoring system 2 repeats the above step S45, and the road monitoring system 2 sends a warning instruction to the next intersection (e.g. Figure 3The continuous intersection warning module 125 then repeats steps S46 to S48, and the steps S45 to S48 are repeated until the result of step S47 is negative, i.e., the continuous intersection warning module 125 determines that the repeated factors and the factor combination no longer include the high-risk driving factor (i.e., the speeding mentioned above) generated by the vehicle 4, or that although the repeated factors and the factor combination still include the high-risk driving factor generated by the vehicle 4, the similarity value of the factor combination is lower than the first default value, and the process enters Figure 5 Node A and end the monitoring process of vehicle 4.

[0049] Or, as Figure 5 In steps S49 and S50, the continuous intersection warning module 125 starts to accumulate the number of times N (N=1) and determines whether the accumulated number of times N reaches a default number, such as 2. If not (N is less than 2), Figure 5 In step S51, the continuous intersection warning module 125 generates a monitoring message and provides it to the road monitoring system 2. The monitoring message includes the traveling direction of the vehicle 4. The default number of times can also be set to any other positive integer.

[0050] Then, if Figure 5 In step S52, the road monitoring system sends a monitoring instruction to the next intersection (e.g., the intersection) that the vehicle 4 is about to arrive at or may arrive at according to the direction of travel of the vehicle included in the warning message provided in step S51. Figure 3 A monitor 36 is provided at the intersection D) shown in the figure, so that the monitor 36 at the next intersection (intersection D) that the vehicle 4 is about to arrive at or may arrive at captures an intersection video and provides it to the computer device 1.

[0051] Then, repeat the above steps S46 to S47. If the result of the judgment in step S47 is no, and the cumulative number N reaches 2 in step S49, and in step S50, it is judged that N=2, then enter the process of Figure 5 Node A of the vehicle 4 is reached and the monitoring process for the vehicle 4 ends. On the other hand, if the result of the judgment in step S47 is yes (indicating that the vehicle 4 is still speeding at intersection D), the continuous intersection warning module 125 resets the accumulated number N to zero and repeats steps S48, S45 to S47 to continuously monitor each intersection that the vehicle 4 may pass through next and issue a warning in advance.

[0052] Returning to step S42, when the judgment result of step S42 is negative, it means that the similarity value generated by comparing the multiple factor combinations appearing in the intersection video with the factor combinations of vehicle-to-pedestrian accidents (or other types of accidents) is less than the similarity threshold and also less than the first default value, indicating that the intersection environment does not have the conditions for the driving of vehicle 4 to cause major casualties. At this time, other reporting and penalty measures should be used to curb such behavior, and monitoring and warnings related to vehicle 4 can be stopped.

[0053] Returning to step S43 , when the determination result of step S43 is negative, it indicates that the vehicle 4 has not exhibited any high-risk driving behavior, and the monitoring and warning related to the vehicle 4 can be stopped.

[0054] In summary, the above embodiment utilizes the accident classification module 121 based on an artificial intelligence algorithm to determine whether an accident has occurred in a traffic scene video and classify traffic scene videos in which accidents have occurred. Furthermore, the multiple accident factor discovery modules identify and statistically analyze all factors that lead to accidents in accident videos of different categories. The accident summary module 125 then summarizes the correlations between certain factors and various accidents based on the statistically analyzed factors. In addition to being able to determine whether the probability of an accident occurring on a road section shown in a newly input traffic scene video is high based on the correlations between certain factors and various accidents, the system can also use various factors appearing in the newly input traffic scene video to determine whether the probability of an accident occurring is high, thereby issuing an early warning to eliminate or reduce the factors that may cause an accident. Furthermore, the system can automatically and quickly review a large number of traffic scene videos and identify possible factors that may have led to an accident. This not only reduces the burden on personnel involved in searching for accident-related factors from a large number of videos and increases their efficiency, but also makes it easier to find certain minor factors or details in the videos that may be related to the accident, thereby achieving the effectiveness and purpose of the present invention.

[0055] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for predicting traffic accidents, characterized in that: include: (A) When an accident classification module of a computer device pre-stored with a large number of traffic scene videos determines that a collision has occurred in the traffic scene videos and confirms the existence of an accident, the traffic scene videos in which the accident occurred are classified into at least three types of accident videos, namely, vehicle-to-pedestrian, vehicle-to-vehicle, and self-collision, based on the objects of the collision, and the accident classification module stores the data in a storage unit of the computer device; the accident classification module obtains first factor data from each type of accident video within a fixed time period from a time point of the accident occurrence to a time point before the time point and records the data in the storage unit; the first factor data includes vehicles, pedestrians, and traffic signal equipment; (B) an accident factor discovery module of the computer device reads the various accident videos from the storage unit and obtains, from the various accident videos, second factor data other than the first factor data within a fixed time period from the time point at which the accident occurred to the time point before the accident; the second factor data includes geographic information and weather information; and (C) An accident summarization module of the computer device integrates all factors of the accident environment in each accident video based on the first factor data and the second factor data obtained by the accident classification module and the accident factor discovery module, and further summarizes various factor combinations in the same accident video, and generates a report based on the above-mentioned summarized various factor combinations, which presents the various factor combinations of the various accident videos occurring at different road sections or intersections.

2. The method for predicting traffic accidents according to claim 1, characterized in that: In step (A), the accident classification module also determines that when a vehicle in one of the traffic scene videos suddenly brakes or turns, it is determined that a similar accident has occurred in the traffic scene video, and classifies the traffic scene video in which the similar accident has occurred as a similar accident video; and in step (B), the accident factor discovery module also summarizes the similar accident videos, and finds the second factor data in the accident environment in addition to the first factor data from the similar accident video; and the various accidents in step (C) also include similar accidents.

3. The method for predicting traffic accidents according to claim 1, characterized in that: Also includes: An accident prediction module of the computer device detects multiple factors appearing in an input traffic scene video to be predicted, and compares the factors in the traffic scene video to be predicted with the factor combinations corresponding to the various accidents summarized in the report to generate a similarity value. When it is determined that the similarity value meets a similarity threshold, an accident warning notification is generated and output.

4. The method for predicting traffic accidents according to claim 3, characterized in that: Also includes: (D) A road monitoring system receives a video of an intersection captured by a monitor installed at the intersection and provides the video to the computer device. The accident prediction module of the computer device detects multiple factors appearing in the intersection video and compares the factors in the intersection video with combinations of factors corresponding to various accidents to generate a similarity value. When the accident prediction module determines that the similarity value associated with the intersection video is less than a similarity threshold but greater than a first default value, a continuous intersection warning module is activated. The continuous intersection warning module determines whether a portion of a factor combination in the intersection video that overlaps with a factor combination corresponding to various accidents includes a high-risk driving factor associated with a vehicle passing through the intersection. If so, the continuous intersection warning module generates a warning message and provides the warning message to the road monitoring system. The warning message includes the direction of travel of the vehicle. (E) after receiving the warning message, the road monitoring system transmits a warning instruction to a warning device located at the next intersection that the vehicle is about to arrive at based on the direction of travel of the vehicle contained in the warning message, causing the warning device to output a warning message based on the warning instruction, and transmits a monitoring instruction to a monitor located at the next intersection, causing the monitor at the next intersection to capture a video of the intersection and provide it to the computer device; (F) when the continuous intersection warning module of the computer device determines that the similarity value associated with the intersection video of step (E) still remains less than the similarity threshold but greater than the first default value, and a factor combination of the intersection video of step (E) and the overlapping portion of each factor combination corresponding to various accidents still include the high-risk driving factor associated with the vehicle, the continuous intersection warning module generates a warning message and provides it to the road monitoring system, wherein the warning message includes the direction of travel of the vehicle; and (G) Repeat steps (E) and (F) until the continuous intersection warning module determines that the similarity value associated with the intersection video in step (E) is less than the first default value, or the repeated part of the factor combination of the intersection video in step (E) and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle.

5. The method for predicting traffic accidents according to claim 4, characterized in that: In step (G), if the similarity value associated with the intersection video analyzed by the continuous intersection warning module in step (E) is less than the first default value, or if the overlapping portion of the factor combination in the intersection video in step (E) and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle, the continuous intersection warning module generates a monitoring message and provides it to the road monitoring system, the monitoring message including the direction of travel of the vehicle; and the method further includes the following steps after step (G): (H) the road monitoring system transmits a monitoring instruction to a monitor located at the next intersection that the vehicle is about to arrive at based on the vehicle's traveling direction included in the monitoring information in step (G), causing the monitor at the next intersection that the vehicle is about to arrive at to capture a video of the intersection and provide it to the computer device; and (I) Repeat steps (G) and (H) until the continuous intersection warning module determines that the similarity value associated with N consecutive (N≥2) intersection videos is less than the first default value, or the repeated portion of the factor combination of the N consecutive intersection videos and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle.

6. The method for predicting traffic accidents according to claim 3, characterized in that: The accident warning notification is also transmitted via Internet of Vehicles related equipment to the on-board device of the vehicle that is about to pass through the traffic section shown in the traffic scene video to be predicted.

7. A computer device for predicting traffic accidents, characterized in that: include: a storage unit in which a large number of traffic scene videos are pre-stored; and A processing unit is electrically connected to the storage unit to access the storage unit, and the processing unit includes an accident classification module, an accident factor discovery module and an accident summary module; wherein When the accident classification module determines that a collision has occurred in the traffic scene videos and confirms the existence of an accident, it classifies the traffic scene videos in which the accident occurred into at least three types of accident videos, namely, vehicle-to-pedestrian, vehicle-to-vehicle, and self-collision, based on the objects of the collision, and stores the classifications in the storage unit. Furthermore, the classification module obtains first factor data from each type of accident video within a fixed time period from a time point at which the accident occurred to a time point before the time point, and records the data in the storage unit. The first factor data includes vehicle, pedestrian, and traffic signal equipment. The accident factor discovery module reads the various accident videos from the storage unit and obtains second factor data, other than the first factor data, from the various accident videos within a fixed time period from the time point of the accident occurrence to the time point before the accident occurrence; the second factor data includes geographic information and weather information; The accident summary module integrates all factors of the accident environment in each accident video based on the first factor data and the second factor data obtained by the accident classification module and the accident factor discovery module, and further summarizes various factor combinations in the same accident video, and generates a report based on the above-mentioned summarized various factor combination results. The report presents the various factor combinations of the various accident videos occurring at different road sections or intersections.

8. The computer device for predicting traffic accidents according to claim 7, characterized in that: The accident classification module also determines that when a vehicle in one of the traffic scene videos suddenly brakes or turns, it determines that a similar accident has occurred in the traffic scene video, and classifies the traffic scene video in which the similar accident has occurred as a similar accident video; and the accident factor discovery module also summarizes the similar accident videos, and finds the second factor data in the accident environment in addition to the first factor data from the similar accident video; and the various accidents also include similar accidents.

9. The computer device for predicting traffic accidents according to claim 7, wherein: The processing unit also includes an accident prediction module, which detects multiple factors appearing in an input traffic scene video to be predicted, and compares the factors in the traffic scene video to be predicted with the factor combinations corresponding to the various accidents summarized in the report to generate a similarity value. When it is determined that the similarity value meets a similarity threshold, an accident warning notification is generated and output.

10. The computer device for predicting traffic accidents according to claim 9, characterized in that: The computer device is also electrically connected to a road monitoring system, and the processing unit further includes a continuous intersection warning module; and the computer device and the road monitoring system perform the following steps: (A) The road monitoring system receives a video of an intersection captured by a monitor installed at the intersection and provides the video to the computer device. The accident prediction module of the computer device detects multiple factors appearing in the intersection video and compares the factors in the intersection video with combinations of factors corresponding to various accidents to generate a similarity value. When the accident prediction module determines that the similarity value associated with the intersection video is less than a similarity threshold but greater than a first default value, the continuous intersection warning module is activated. The continuous intersection warning module determines whether a portion of a factor combination in the intersection video that overlaps with a factor combination corresponding to various accidents includes a high-risk driving factor associated with a vehicle passing through the intersection. If so, the continuous intersection warning module generates a warning message and provides the warning message to the road monitoring system. The warning message includes the direction of travel of the vehicle. (B) after receiving the warning message, the road monitoring system transmits a warning instruction to a warning device located at the next intersection that the vehicle is about to arrive at based on the direction of travel of the vehicle contained in the warning message, causing the warning device to output a warning message based on the warning instruction, and transmits a monitoring instruction to a monitor located at the next intersection, causing the monitor at the next intersection to capture a video of the intersection and provide it to the computer device; (C) when the continuous intersection warning module of the computer device determines that the similarity value associated with the intersection video in step (B) remains less than the similarity threshold but greater than the first default value, and a factor combination in the intersection video in step (B) overlaps with each factor combination corresponding to various accidents and still includes the high-risk driving factor associated with the vehicle, the continuous intersection warning module generates a warning message and provides it to the road monitoring system, the warning message including the direction of travel of the vehicle; and (D) Repeat steps (B) and (C) until the continuous intersection warning module determines that the similarity value associated with the intersection video of step (E) is less than the first default value, or the repeated part of the factor combination of the intersection video of step (E) and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle.

11. The computer device for predicting traffic accidents according to claim 10, wherein: In step (D), if the similarity value associated with the intersection video analyzed by the continuous intersection warning module in step (B) is less than the first default value, or if the overlapping portion of the factor combination in the intersection video in step (B) and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle, the continuous intersection warning module generates a monitoring message and provides it to the road monitoring system, the monitoring message including the vehicle's direction of travel; and the computer device and the road monitoring system further perform the following steps after step (D): (E) the road monitoring system transmits a monitoring instruction to a monitor located at the next intersection that the vehicle is about to arrive at based on the vehicle's travel direction included in the monitoring information in step (D), causing the monitor at the next intersection that the vehicle is about to arrive at to capture a video of the intersection and provide it to the computer device; and (F) Repeat steps (D) and (E) until the continuous intersection warning module determines that the similarity value associated with N consecutive (N≥2) intersection videos is less than the first default value, or the repeated part of the factor combination of the N consecutive intersection videos and the factor combinations corresponding to various accidents does not include the high-risk driving factor associated with the vehicle.

12. The computer device for predicting traffic accidents according to claim 9, characterized in that: The accident warning notification is also transmitted via Internet of Vehicles related equipment to the on-board device of the vehicle that is about to pass through the traffic section shown in the traffic scene video to be predicted.

13. A computer-readable recording medium for predicting traffic accidents, characterized in that: A software program including an accident classification module, an accident factor discovery module, an accident summarization module, an accident prediction module and a continuous intersection warning module is stored therein. When the software program is loaded and executed by a computer device pre-stored with a large number of traffic scene videos, the computer device can complete the method for predicting traffic accidents according to any one of claims 1 to 6.

Images