Vehicle collision detection methods, devices, equipment and storage media

Abstract

This application provides a vehicle collision detection method, apparatus, device, and storage medium, relating to the field of image processing technology. The method includes: during vehicle collision detection, acquiring a current frame image using an image acquisition device in the vehicle; determining a corresponding target difference image based on the current frame image and at least one previous frame image, wherein the acquisition time of the previous at least one frame image and the acquisition time of the current frame image are within a preset time period; then converting the target difference image from the time domain to the frequency domain, and determining the vehicle collision detection result based on the high-frequency components in the conversion result. Using the technical solution provided in this application, vehicle collision detection is achieved without adding a gravity sensor to the vehicle.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a vehicle collision detection method, apparatus, device, and storage medium. Background Technology

[0002] With the increasing number of cars on the road, traffic accidents and incidents are also on the rise. Dashcams can preserve evidence of unexpected situations that may occur to a vehicle, thus protecting the legal rights of drivers.

[0003] Given the limited memory card capacity of dashcams, collision detection is necessary to store targeted footage related to vehicle collisions during driving. Currently, this is primarily achieved through gravity sensors, but this requires the installation of additional gravity sensors within the vehicle.

[0004] Therefore, how to perform vehicle collision detection without adding an additional gravity sensor to the vehicle is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This application provides a vehicle collision detection method, apparatus, device, and storage medium to solve the technical problem of how to perform vehicle collision detection without adding an additional gravity sensor to the vehicle in the prior art.

[0006] This application provides a vehicle collision detection method, including: The current frame image is captured using the image acquisition device in the vehicle; Based on the current frame image and at least one previous frame image, a corresponding target difference image is determined; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period; The target differential image is converted from the time domain to the frequency domain, and the collision detection result of the vehicle is determined based on the high-frequency components in the conversion result.

[0007] According to a vehicle collision detection method provided in this application, determining the vehicle collision detection result based on the high-frequency components in the conversion result includes: If the high-frequency component is greater than a preset threshold, the collision detection result is determined to be a collision.

[0008] According to a vehicle collision detection method provided in this application, the step of determining that a collision has occurred when the high-frequency component is greater than a preset threshold includes: When the high-frequency component is greater than a preset threshold, the actual driving speed of the vehicle and the driving speed detected by radar are obtained; Based on the actual driving speed and the driving speed detected by the radar, the probability of the vehicle colliding is determined; If the probability is greater than the probability threshold, the collision detection result is determined to be a collision.

[0009] According to a vehicle collision detection method provided in this application, determining the corresponding target difference image based on the current frame image and at least one previous frame image includes: When the number of the preceding at least one frame image is multiple frames, the difference images corresponding to the current frame image and each frame image are determined respectively to obtain multiple frame difference images; The target difference image is obtained by weighted summation of the multiple difference images.

[0010] According to the vehicle collision detection method provided in this application, the method further includes: When the collision detection result indicates that the vehicle has experienced a collision event, a target video block containing the video of the collision event is determined from multiple video blocks divided in the vehicle's memory, based on the time at which the collision detection result is determined. Based on the type of the collision event, determine the status information corresponding to the target video block, the status information including full coverage priority and / or locking information; The type of the collision event and the status information are used to determine the index information of the target video block, and the index information is stored.

[0011] According to the vehicle collision detection method provided in this application, determining the state information corresponding to the target video block based on the type of the collision event includes: When the collision event includes at least two types, the state information corresponding to each type of event is determined respectively; If the status information includes full coverage priority, the lowest full coverage priority is determined as the full coverage priority corresponding to the target video block; If the status information includes locking information, and there exists a locking status corresponding to at least one type of event, then the locking status is determined as the locking information corresponding to the target video block.

[0012] According to a vehicle collision detection method provided in this application, storing the index information includes: When the target video block includes an in-vehicle video block and an out-of-vehicle video block, the index information corresponding to the in-vehicle video block is stored in the index information corresponding to the out-of-vehicle video block; The index information corresponding to the external video recording block is stored in the index information corresponding to the internal video recording block.

[0013] This application also provides a vehicle collision detection device, including: The acquisition unit is used to acquire the current frame image through the image acquisition device in the vehicle; The first processing unit is configured to determine a corresponding target differential image based on the current frame image and at least one previous frame image; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period. The second processing unit is used to convert the target differential image from the time domain to the frequency domain, and determine the collision detection result of the vehicle based on the high-frequency components in the conversion result.

[0014] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the vehicle collision detection method as described above.

[0015] This application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle collision detection method as described above.

[0016] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle collision detection method as described above.

[0017] The vehicle collision detection method, apparatus, device, and storage medium provided in this application, when performing vehicle collision detection, can acquire the current frame image through an image acquisition device in the vehicle; and determine the corresponding target difference image based on the current frame image and at least one previous frame image, wherein the acquisition time of the previous at least one frame image and the acquisition time of the current frame image are within a preset time period; then, the target difference image is converted from the time domain to the frequency domain, and the vehicle collision detection result is determined based on the high-frequency components in the conversion result. In this way, by determining the target difference image corresponding to at least two consecutive frames, converting the target difference image from the time domain to the frequency domain, and determining the vehicle collision detection result based on the high-frequency components in the conversion result, there is no need to add an additional gravity sensor to the vehicle, thus achieving vehicle collision detection without adding a gravity sensor to the vehicle. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a vehicle collision detection method provided in an embodiment of this application.

[0020] Figure 2 This is a flowchart illustrating a video block storage method provided in an embodiment of this application.

[0021] Figure 3 This is a schematic diagram of the structure of a first-level index file format provided in an embodiment of this application.

[0022] Figure 4 This is a schematic diagram of the structure of a vehicle collision detection device provided in an embodiment of this application.

[0023] Figure 5 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] In the embodiments of this application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone, where A and B can be singular or plural. In the textual description of this application, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0026] The technical solution provided in this application can be adapted to vehicle collision detection scenarios. Currently, during vehicle operation, collisions are mainly detected using gravity sensors. Specifically, when a vehicle collides, a drastic change in acceleration occurs. The sensing elements inside the gravity sensor, such as a mass block and spring system, respond to the acceleration change, causing the mass block to displace, which in turn causes the spring to extend or contract or its resistance / charge to change. The circuit system converts these physical changes, such as changes in resistance or charge, into electrical signals.

[0027] After the converted electrical signal is obtained, the processor or controller in the vehicle will analyze the electrical signal to determine the degree of change in acceleration. If the change in acceleration exceeds the set threshold and meets the specific collision pattern, such as a sudden and brief acceleration peak, it will be identified as a collision event, that is, a collision has been detected.

[0028] However, the above method requires the addition of a gravity sensor in the vehicle. To achieve vehicle collision detection without adding a gravity sensor, this application provides a vehicle collision detection method. The executing entity can be an in-vehicle terminal, an electronic device such as a server, or a vehicle collision detection device within an electronic device. This vehicle collision detection device can be implemented through software, hardware, or a combination of both.

[0029] The vehicle collision detection method provided in this application will be described in detail below through several specific embodiments. It is understood that these specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0030] Figure 1 This is a schematic flowchart of a vehicle collision detection method provided in an embodiment of this application. For example, please refer to [link to relevant documentation]. Figure 1 As shown, the vehicle collision detection method may include: S101. Acquire the current frame image using the image acquisition device in the vehicle.

[0031] For example, the image acquisition device could be a dashcam.

[0032] S102. Based on the current frame image and at least one previous frame image, determine the corresponding target difference image.

[0033] Among them, the acquisition time of at least the previous frame and the acquisition time of the current frame are within a preset time period.

[0034] For example, in the embodiments of this application, the number of at least one frame image can be one frame, two frames, or other numbers, such as three or four frames, etc., which can be set according to actual needs. The preset time period can be 0.5 seconds, 1 second, etc., which can be set according to actual needs.

[0035] By defining a target difference image, motion information can be captured more effectively. Typically, the difference image highlights moving objects.

[0036] In this application embodiment, when determining the corresponding target difference image based on the current frame image and at least one previous frame image, at least two of the following possible implementation methods may be included: In one possible implementation, the number of the previous at least one frame image is one frame, that is, the corresponding target difference image is determined based on the current frame image and the previous frame image.

[0037] For example, when determining the corresponding target difference image based on the current frame image and the previous frame image, the difference image corresponding to the current frame image and the previous frame image can be determined and the difference image can be determined as the target difference image.

[0038] For example, in the embodiments of this application, when determining the corresponding difference image based on the current frame image and the previous frame image, refer to the following formula 1: Formula 1 in, This represents the difference image between the current frame and the previous frame. Indicates the first The current frame image, Indicates the first The previous frame image of the current frame image and This represents two consecutive frames of images.

[0039] For example, in the embodiments of this application, when determining the corresponding difference image based on the current frame image and the previous frame image, the current frame image and the previous frame image can be preprocessed first, for example, the image can be grayscaled to reduce computational complexity and remove unnecessary information; and the corresponding difference image can be determined based on the preprocessed current frame image and the previous frame image, which can be set according to actual needs.

[0040] In one possible implementation, the number of previous at least one frames is multiple frames, that is, the corresponding target difference image is determined based on the current frame image and the previous multiple frames images.

[0041] For example, when determining the corresponding target difference image based on the current frame image and the previous multiple frames, the difference images corresponding to the current frame image and each of the previous frames can be determined separately to obtain multiple frame difference images; the multiple frame difference images are then weighted and summed to obtain the target difference image.

[0042] It is understood that the specific implementation of determining the difference image corresponding to the current frame image and each frame image is similar to the specific implementation of determining the difference image corresponding to the current frame image and the previous frame image described above. Please refer to the relevant description of determining the difference image corresponding to the current frame image and the previous frame image described above. Here, the embodiments of this application will not be repeated.

[0043] When performing a weighted summation of multiple difference images, the weights for each difference image can be determined based on factors such as the quality of the difference images, the time interval, and the stability of vehicle motion. Generally, frames closer to the current frame are given greater weight to ensure background stability and accurate target detection.

[0044] After determining the target differential image corresponding to at least two consecutive frames, considering that under normal circumstances, when a vehicle is driving normally on the road, the images recorded by the dashcam are relatively stable, the motion information of the target differential image corresponding to at least two consecutive frames is relatively predictable, and the vibration frequency of the vehicle is relatively low and stable; however, when a car accident collision occurs, the vehicle will produce unusual vibrations, which are usually characterized by a sharp increase in high-frequency components. Therefore, in this embodiment, after obtaining the target differential image corresponding to at least two consecutive frames, the vibration frequency can be analyzed and these high-frequency components can be captured to determine the collision detection result of the vehicle, i.e., the following S103 is executed: S103. Convert the target difference image from the time domain to the frequency domain, and determine the vehicle collision detection result based on the high-frequency components in the conversion result.

[0045] For example, in the embodiments of this application, when converting the target difference image from the time domain to the frequency domain, a Fast Fourier Transform (FFT) can be performed on the target difference images corresponding to at least two consecutive frames, as shown in Formula 2 below. After obtaining the conversion result, the high-frequency components in the conversion result are analyzed. If the high-frequency components increase abnormally, it indicates that a vehicle collision has occurred.

[0046] Formula 2 in, The result of the conversion, i.e., the signal in the frequency domain, also known as the spectrum, represents the intensity and phase information of the signal at different frequency components; This represents the value of the target difference image corresponding to the current frame image and at least the previous frame image in the time domain, also known as the time domain signal. It can be a continuous signal or a discrete signal, and is usually represented as a function or sequence. This indicates the application of a window function, used in Fourier transform to limit the time range of a signal. The window function is typically 1 in the time domain, but in some cases other functions can be used to limit the time range of the signal to reduce edge effects and improve the readability of the spectrum. This represents a difference image with a window function applied. Let represent an exponential function, where Represents the imaginary unit. Indicates frequency, It represents time and is used to describe the phase information of a signal in the frequency domain.

[0047] As shown in Formula 2 above, after converting the target difference image from the time domain to the frequency domain and obtaining the conversion result, the high-frequency components in the conversion result can be analyzed, and the vehicle collision detection result can be determined based on the high-frequency components in the conversion result, thereby realizing vehicle collision detection without adding a gravity sensor to the vehicle.

[0048] As can be seen from the embodiments of this application, when performing vehicle collision detection, the current frame image can be acquired through the image acquisition device in the vehicle; and based on the current frame image and at least one previous frame image, the corresponding target difference image is determined; then, the target difference image is converted from the time domain to the frequency domain, and the vehicle collision detection result is determined based on the high-frequency components in the conversion result. In this way, by determining the target difference image corresponding to at least two consecutive frames, converting the target difference image from the time domain to the frequency domain, and determining the vehicle collision detection result based on the high-frequency components in the conversion result, there is no need to add an additional gravity sensor to the vehicle, thus realizing vehicle collision detection without adding a gravity sensor to the vehicle.

[0049] Based on the above Figure 1 In the embodiments shown, when determining the collision detection result of a vehicle based on the high-frequency components in the conversion result, it can be determined whether the high-frequency components are greater than a preset threshold. If the high-frequency components are less than or equal to the preset threshold, the collision detection result is determined to be no collision; conversely, if the high-frequency components are greater than the preset threshold, the collision detection result is determined to be a collision. Thus, by analyzing the high-frequency components in the conversion result, the collision detection result of the vehicle can be obtained.

[0050] The preset threshold value can be determined based on experience or through a training dataset, and can be set according to actual needs.

[0051] For example, in this embodiment of the application, when analyzing the high-frequency components in the conversion result to obtain the vehicle collision detection result, in order to further improve the accuracy of the determined collision detection result, the actual driving speed of the vehicle and the driving speed detected by radar can also be analyzed together to obtain the vehicle collision detection result.

[0052] For example, when combining the vehicle's actual speed and the speed detected by radar to obtain the vehicle's collision detection result, the actual speed and the speed detected by radar can be obtained if the high-frequency component is greater than a preset threshold. Based on the actual speed and the speed detected by radar, the probability of a collision is determined. If the probability is greater than a probability threshold, the collision detection result is determined to be a collision. In this way, when the high-frequency component is greater than the preset threshold, combining the vehicle's actual speed and the speed detected by radar to obtain the vehicle's collision detection result can effectively improve the accuracy of the determined collision detection result.

[0053] For example, in this embodiment of the application, when determining the probability of a vehicle collision based on the actual driving speed and the driving speed detected by radar, the following formula 3 can be used: Formula 3 in, This indicates the probability of a vehicle collision. Indicates the actual speed of the vehicle. This indicates the speed detected by radar. This indicates the adjustment parameter.

[0054] Typically, after a vehicle collision or other accident, if a dashcam can preserve evidence of the incident, it can effectively protect the driver's legal rights. However, considering the limited capacity of a dashcam's memory card, if the footage of the collision or other accident is retrieved from the memory card only after a long interval, the footage may be overwritten. Therefore, in this embodiment, to prevent the footage of the collision or other accident from being overwritten, the footage of the collision can be selectively locked and stored to extend its retention period, thereby providing a basis for subsequent retrieval of the footage. See below for details. Figure 2 The example shown.

[0055] Figure 2 This is a flowchart illustrating a video block storage method provided in an embodiment of this application. For example, please refer to [link to relevant documentation]. Figure 2 As shown, the method may include: S201. When the collision detection result indicates that a collision event has occurred, based on the determined time of the collision detection result, a target video block including the video of the collision event is determined from multiple video blocks divided in the vehicle memory.

[0056] For example, in this embodiment of the application, the memory card storage capacity information can be obtained in advance, and a video superblock and a first-level index file can be generated based on the dashcam version and the memory card storage capacity information. The video superblock and the first-level index file can be a reserved block, and their capacity can be the capacity of a single video block.

[0057] For example, the format of a first-level index file can be found in [reference needed]. Figure 3 As shown, Figure 3 This is a schematic diagram of a first-level index file format provided in an embodiment of this application. In this format, a digest group is equal to a single video file description (128B) * the number of video files + the digest group header, totaling 4KB. Multiple digest groups are divided according to their size, and multiple digest groups are in the same first-level index file.

[0058] When a dashcam records video, it generates video blocks. Upon detecting a collision, based on the determined time of the collision detection result, it identifies a target video block containing the video of the collision event from multiple video blocks allocated in the vehicle's memory. For example, based on the determined time of the collision detection result, it can extend forward by a first duration and backward by a second duration to select a target video block containing the video of the collision event from multiple video blocks.

[0059] The values ​​of the first duration and the second duration can be set according to actual needs. For example, the first duration can be 10 seconds and the second duration can be 60 seconds.

[0060] It is understood that, in the embodiments of this application, in addition to determining the target video block containing the collision event from multiple video blocks and storing the collision event information when a vehicle collision event is detected, the target video block containing the video of other events can also be determined from multiple video blocks and stored when other events are detected, such as airbag deployment, collision event, vehicle stalling, emergency braking, pedestrian detection, unfastened seat belt, and traffic light detection. The specific settings can be configured according to actual needs.

[0061] After the target video block is identified, in order to effectively prevent the target video block from being overwritten and thus better store the target video block, the following step S202 can be executed: S202. Based on the type of collision event, determine the status information corresponding to the target video block. The status information includes full coverage priority and / or locking information.

[0062] The collision event type can include only the collision event type, or it can include other types, such as at least one of the following: airbag deployment type, vehicle stall type, emergency braking type, pedestrian detection type, unfastened seat belt type, or traffic light detection type. The specific type can be set according to actual needs.

[0063] For example, if the collision event type includes both collision event type and unfastened seat belt type, it means that the vehicle not only collided, but the driver was not wearing a seat belt; or if the collision event type includes both collision event type and emergency braking type, it means that the vehicle not only collided, but also performed an emergency braking operation.

[0064] Among them, full coverage priority refers to the priority level of a video block being overwritten when multiple video blocks are already full. The higher the full coverage priority, the higher the priority level of the video block being overwritten; conversely, the lower the full coverage priority, the lower the priority level of the video block being overwritten.

[0065] The locking information can include a locked status or an unlocked status. If the locking information is "locked," the recording is less likely to be overwritten when multiple recording blocks are full; if the locking information is "unlocked," the recording is more likely to be overwritten when multiple recording blocks are full.

[0066] For example, in the embodiments of this application, when determining the status information corresponding to the target video block based on the type of collision event, if the collision event includes at least two types, the status information corresponding to each type of event can be determined separately; if the status information includes full coverage priority, the lowest full coverage priority can be determined as the full coverage priority corresponding to the target video block; if the status information includes locking information, if there is a locking status corresponding to at least one type of event, the locking status can be determined as the locking information corresponding to the target video block, thereby determining the status information corresponding to the target video block.

[0067] For example, consider a target video block containing a video of a collision event A. The collision event A can be categorized into two types: collision event type and unfastened seatbelt type. When determining the full coverage priority of the target video block, considering that the full coverage priority of the collision event type is generally lower than that of the unfastened seatbelt type, the full coverage priority of the collision event type can be used to determine the full coverage priority of the target video block. This makes it less likely that the video of collision event A will be overwritten when multiple video blocks are already full, thus allowing for better storage of the target video block. Similarly, when determining the locking information of the target video block, considering that the collision event type is an important event, the locking information for the collision event is a state lock. The unfastened seatbelt type is a relatively unimportant event, so the state lock can be directly determined as the locking information for the target video block. This makes it less likely that the video of collision event A will be overwritten when multiple video blocks are already full, thus allowing for better storage of the target video block.

[0068] After determining the type and status information of the collision event, the following step S203 can be executed: the type and status information of the collision event are determined as the index information of the target video block. In this embodiment, this can be recorded as secondary index information and stored. For example, the type, full coverage priority, and status information of the collision events included in the video block can also be updated to the summary information of the primary index file, and the information of the video superblock can be updated according to the primary index file. Specific settings can be configured according to actual needs.

[0069] S203. Determine the type and status information of the collision event as the index information of the target video block, and store the index information.

[0070] When storing index information, considering that during vehicle operation, the dual-camera dashcam records exterior view R1 and interior view R2 respectively, the exterior view is stored normally, while the interior view is only stored in the event of a collision. Furthermore, the interior view is stored in separate video blocks, with each storage session lasting t seconds. If no collision occurs within those t seconds, the next storage session starts overwriting the previous one; if a collision occurs within those t seconds, the next storage session begins from the t-second mark to update the video block.

[0071] Furthermore, to facilitate the retrieval of in-vehicle and out-of-vehicle footage of collision events, for example, in this embodiment of the application, when storing index information, if the target video block includes both in-vehicle and out-of-vehicle video blocks, the index information corresponding to the in-vehicle video block can be stored in the index information corresponding to the out-of-vehicle video block. For example, the index information corresponding to the in-vehicle video block can be linked to the index information corresponding to the out-of-vehicle video block. In this way, the in-vehicle video block can be found simultaneously through the index information corresponding to the in-vehicle video block linked to the index information corresponding to the out-of-vehicle video block. Alternatively, the index information corresponding to the out-of-vehicle video block can be stored in the index information corresponding to the in-vehicle video block. For example, the index information corresponding to the out-of-vehicle video block can be linked to the index information corresponding to the in-vehicle video block. In this way, the out-of-vehicle video block can be found simultaneously through the index information corresponding to the out-of-vehicle video block linked to the index information corresponding to the in-vehicle video block. This facilitates the simultaneous retrieval of both in-vehicle and out-of-vehicle video blocks, providing convenience for users to find collision events.

[0072] As can be seen, in this embodiment of the application, when the collision detection result indicates that a vehicle has experienced a collision event, a target video block containing the video of the collision event can be determined from multiple video blocks divided in the vehicle's memory based on the determined time of the collision detection result; and based on the type of the collision event, the corresponding status information of the target video block is determined, including full coverage priority and / or locking information; the type and status information of the collision event are determined as the index information of the target video block, and the index information is stored. In this way, storing the type and status information of the collision event as the index information of the target video block allows for more targeted storage of video blocks without the need for additional memory cards, making more efficient use of storage space and providing convenience for users to find collision events.

[0073] For example, based on the above Figure 2 In the illustrated embodiment, when multiple video blocks are full, the video block corresponding to the highest full coverage priority is determined as an overwhelmable video block; and / or, the video block corresponding to the unlocked state is determined as an overwhelmable video block; the newly acquired image by the image acquisition device is stored in the overwhelmable video block. This effectively avoids the video blocks corresponding to low full coverage priority being overwritten, and / or the video blocks with locked state being overwritten, extending the retention period of collision events, thereby providing a basis for recording subsequent vehicle collisions and other accidents.

[0074] The vehicle collision detection device provided in this application is described below. The vehicle collision detection device described below can be referred to in correspondence with the vehicle collision detection method described above.

[0075] Figure 4 This is a schematic diagram of the structure of a vehicle collision detection device provided in an embodiment of this application. For example, please refer to [link to relevant documentation]. Figure 4 As shown, the vehicle collision detection device 40 may include: Acquisition unit 401 is used to acquire the current frame image through the image acquisition device in the vehicle; The first processing unit 402 is used to determine a corresponding target difference image based on the current frame image and the previous at least one frame image; wherein the acquisition time of the previous at least one frame image and the acquisition time of the current frame image are within a preset time period. The second processing unit 403 is used to convert the target differential image from the time domain to the frequency domain, and determine the collision detection result of the vehicle based on the high-frequency components in the conversion result.

[0076] For example, in an embodiment of this application, the second processing unit 403 is used to determine the collision detection result of the vehicle based on the high-frequency components in the conversion result, including: If the high-frequency component is greater than a preset threshold, the collision detection result is determined to be a collision.

[0077] For example, in an embodiment of this application, the second processing unit 403 is configured to determine that a collision has occurred when the high-frequency component is greater than a preset threshold, including: When the high-frequency component is greater than a preset threshold, the actual driving speed of the vehicle and the driving speed detected by radar are obtained; Based on the actual driving speed and the driving speed detected by the radar, the probability of the vehicle colliding is determined; If the probability is greater than the probability threshold, the collision detection result is determined to be a collision.

[0078] For example, in an embodiment of this application, the first processing unit 402 is configured to determine a corresponding target difference image based on the current frame image and at least one previous frame image, including: When the number of the preceding at least one frame image is multiple frames, the difference images corresponding to the current frame image and each frame image are determined respectively to obtain multiple frame difference images; The target difference image is obtained by weighted summation of the multiple difference images.

[0079] For example, in an embodiment of this application, the vehicle collision detection device 40 further includes: The third processing unit is used to determine, based on the time of determination of the collision detection result, a target video block including the video of the collision event from multiple video blocks divided in the vehicle memory when the collision detection result indicates that the vehicle has experienced a collision event. The fourth processing unit is used to determine the status information corresponding to the target video block based on the type of the collision event, wherein the status information includes full coverage priority and / or locking information; The fifth processing unit is used to determine the type of the collision event and the status information as the index information of the target video block, and to store the index information.

[0080] For example, in an embodiment of this application, the fourth processing unit is used to determine the state information corresponding to the target video block based on the type of the collision event, including: When the collision event includes at least two types, the state information corresponding to each type of event is determined respectively; If the status information includes full coverage priority, the lowest full coverage priority is determined as the full coverage priority corresponding to the target video block; If the status information includes locking information, and there exists a locking status corresponding to at least one type of event, then the locking status is determined as the locking information corresponding to the target video block.

[0081] For example, in an embodiment of this application, the fifth processing unit is used to store the index information, including: When the target video block includes an in-vehicle video block and an out-of-vehicle video block, the index information corresponding to the in-vehicle video block is stored in the index information corresponding to the out-of-vehicle video block; The index information corresponding to the external video recording block is stored in the index information corresponding to the internal video recording block.

[0082] The vehicle collision detection device 40 provided in this application embodiment can execute the technical solution of the vehicle collision detection method in any of the above embodiments. Its implementation principle and beneficial effects are similar to those of the vehicle collision detection method. Please refer to the implementation principle and beneficial effects of the vehicle collision detection method. It will not be repeated here.

[0083] Figure 5 This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of this application, such as... Figure 5As shown, the electronic device may include a processor 510, a communications interface 520, a memory 530, and a communication bus 540, wherein the processor 510, communications interface 520, and memory 530 communicate with each other via the communication bus 540. The processor 510 can call logical instructions in the memory 530 to execute a vehicle collision detection method, which includes: acquiring a current frame image using an image acquisition device in the vehicle; determining a corresponding target differential image based on the current frame image and at least one previous frame image; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period; converting the target differential image from the time domain to the frequency domain, and determining the vehicle collision detection result based on the high-frequency components in the conversion result.

[0084] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0085] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the vehicle collision detection method provided by the above methods. The method includes: acquiring a current frame image through an image acquisition device in the vehicle; determining a corresponding target difference image based on the current frame image and at least one previous frame image; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period; converting the target difference image from the time domain to the frequency domain, and determining the vehicle collision detection result based on the high-frequency components in the conversion result.

[0086] In another aspect, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the vehicle collision detection method provided by the above methods. The method includes: acquiring a current frame image through an image acquisition device in a vehicle; determining a corresponding target difference image based on the current frame image and at least one previous frame image; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period; converting the target difference image from the time domain to the frequency domain, and determining the vehicle collision detection result based on the high-frequency components in the conversion result.

[0087] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0088] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A vehicle collision detection method, characterized in that, include: The current frame image is captured using the image acquisition device in the vehicle; Based on the current frame image and at least one previous frame image, a corresponding target difference image is determined; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period; The target differential image is converted from the time domain to the frequency domain, and the collision detection result of the vehicle is determined based on the high-frequency components in the conversion result.

2. The vehicle collision detection method according to claim 1, characterized in that, Determining the collision detection result of the vehicle based on the high-frequency components in the conversion result includes: If the high-frequency component is greater than a preset threshold, the collision detection result is determined to be a collision.

3. The vehicle collision detection method according to claim 2, characterized in that, Determining that a collision has occurred when the high-frequency component is greater than a preset threshold includes: When the high-frequency component is greater than a preset threshold, the actual driving speed of the vehicle and the driving speed detected by radar are obtained; Based on the actual driving speed and the driving speed detected by the radar, the probability of the vehicle colliding is determined; If the probability is greater than the probability threshold, the collision detection result is determined to be a collision.

4. The vehicle collision detection method according to any one of claims 1-3, characterized in that, Determining the corresponding target difference image based on the current frame image and at least one previous frame image includes: When the number of the preceding at least one frame image is multiple frames, the difference images corresponding to the current frame image and each frame image are determined respectively to obtain multiple frame difference images; The target difference image is obtained by weighted summation of the multiple difference images.

5. The vehicle collision detection method according to any one of claims 1-3, characterized in that, The method further includes: When the collision detection result indicates that the vehicle has experienced a collision event, a target video block containing the video of the collision event is determined from multiple video blocks divided in the vehicle's memory, based on the time at which the collision detection result is determined. Based on the type of the collision event, determine the status information corresponding to the target video block, the status information including full coverage priority and / or locking information; The type of the collision event and the status information are used to determine the index information of the target video block, and the index information is stored.

6. The vehicle collision detection method according to claim 5, characterized in that, Determining the status information corresponding to the target video block based on the type of the collision event includes: When the collision event includes at least two types, the state information corresponding to each type of event is determined respectively; If the status information includes full coverage priority, the lowest full coverage priority is determined as the full coverage priority corresponding to the target video block; If the status information includes locking information, and there exists a locking status corresponding to at least one type of event, then the locking status is determined as the locking information corresponding to the target video block.

7. The vehicle collision detection method according to claim 5, characterized in that, The storage of the index information includes: When the target video block includes an in-vehicle video block and an out-of-vehicle video block, the index information corresponding to the in-vehicle video block is stored in the index information corresponding to the out-of-vehicle video block; The index information corresponding to the external video recording block is stored in the index information corresponding to the internal video recording block.

8. A vehicle collision detection device, characterized in that, include: The acquisition unit is used to acquire the current frame image through the image acquisition device in the vehicle; The first processing unit is configured to determine a corresponding target differential image based on the current frame image and at least one previous frame image; wherein the acquisition time of the at least one previous frame image and the acquisition time of the current frame image are within a preset time period. The second processing unit is used to convert the target differential image from the time domain to the frequency domain, and determine the collision detection result of the vehicle based on the high-frequency components in the conversion result.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the vehicle collision detection method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the vehicle collision detection method as described in any one of claims 1 to 7.

Images