Image recognition method and system for assisted driving

Obtaining road and vehicle data through on-board cameras, predicting potential hazard points and performing driving interventions, solving the problem that the existing technology cannot effectively avoid the 'ghost probe', and improving driving safety and experience.

CN118514696BActive Publication Date: 2024-11-22WUXI DINGYI INNOVATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410707599.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-03
Publication Date
2024-11-22
Estimated Expiration
2044-06-03

AI Technical Summary

Technical Problem

The existing driving assistance system cannot effectively predict and avoid the 'ghost probe', resulting in the vehicle's inability to avoid or frequent emergency stop intervention, and the experience is poor.

Method used

The on-board cameras obtain road structure image data and driving vehicle data, predict potential hazard points, and analyze the relationship between driving behavior and specific behaviors, and perform appropriate driving interventions, such as early warning or braking or lane change avoidance instructions.

Benefits of technology

Effectively reduce the impact of the "ghost probe" on vehicle driving safety, improve driving safety, and reduce the intensity of driving intervention through advance prediction and timely intervention, and improve the driving experience of occupants.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118514696B_ABST
    Figure CN118514696B_ABST
Patent Text Reader

Abstract

The present invention belongs to the field of intelligent driving technology. An image recognition method and system for assisted driving are provided. The method comprises: using a vehicle-mounted camera to obtain road structure image data within a first distance range in front of the vehicle, and predicting a number of first potential danger points based on the road structure image data; using a vehicle-mounted camera to obtain first driving vehicle data within a second distance range in front of the vehicle, and predicting the relationship between the driving behavior of the corresponding driving vehicle and the specific behavior based on the first driving vehicle data, and obtaining a number of second potential danger points; performing driving intervention on the vehicle for each first potential danger point and second potential danger point. The present invention uses a vehicle-mounted camera to analyze the road structure and the driving behavior of surrounding vehicles in advance, and takes appropriate driving intervention operations in a timely manner, which effectively reduces the impact of "ghosting" on the driving safety of the vehicle, and can also reduce the intensity of driving intervention, bringing a more comfortable driving experience to the passengers in the vehicle.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to an image recognition method for assisted driving. Background Art

[0002] To improve driving safety, more and more vehicles are equipped with detection sensors to sense the real-time conditions around the vehicle, and make driving interventions or adjustments to driving strategies accordingly. "Ghost peeking" is a behavior that can easily lead to serious traffic accidents, but the existing driving assistance system cannot effectively predict "ghost peeking" and avoid the impact of "ghost peeking", resulting in vehicles either being unable to avoid "ghost peeking" or frequently making emergency stops, resulting in a poor experience. Summary of the invention

[0003] In this regard, the present invention provides an image recognition method, system, electronic device, computer storage medium and computer program product for assisted driving to solve the above-mentioned technical problems.

[0004] The present invention provides an image recognition method for assisted driving, the method comprising the following steps:

[0005] Using a vehicle-mounted camera to obtain road structure image data within a first distance ahead of the vehicle, and predicting a number of first potential danger points based on the road structure image data;

[0006] Using a vehicle-mounted camera to obtain first traveling vehicle data within a second distance range ahead of the vehicle, predicting a relationship between a driving behavior of a corresponding traveling vehicle and a specific behavior based on the first traveling vehicle data, and obtaining a plurality of second potential danger points;

[0007] performing driving intervention on the vehicle at each of the first potential danger point and the second potential danger point;

[0008] Among them, the first potential danger point and the second potential danger point are both related to undetected objects that enter the road where the vehicle is traveling, and the specific behavior is that the undetected objects enter the road where the vehicle is traveling.

[0009] Optionally, the method of using a vehicle-mounted camera to obtain road structure image data within a first distance range ahead of the vehicle includes:

[0010] Using a vehicle-mounted camera to obtain data of a second vehicle in a lane on the right side of the vehicle within a third distance range in front of the vehicle, wherein the second vehicle data includes the number of vehicles, the average speed of the vehicles, and the average distance between the vehicles;

[0011] The first distance range is determined according to the second driving vehicle data.

[0012] Optionally, determining the first distance range according to the second driving vehicle data includes:

[0013] Substituting the number of vehicles, the average speed of the vehicles and the average distance between the vehicles into the following formula for calculation, the first distance range is obtained: ; In the formula, is the first distance range, is the number of vehicles, is the average speed of the vehicle, is the average distance between vehicles, , , is the corresponding dimensionless conversion function.

[0014] Optionally, determining the first distance range according to the second driving vehicle data includes:

[0015] Obtaining a historical lane change probability of each lane in a driving area where the vehicle is located, and correcting the second driving vehicle data according to the historical lane change probability to obtain third driving vehicle data;

[0016] The first distance range is determined according to the third vehicle driving data.

[0017] Optionally, the first driving vehicle data includes deceleration data, a vehicle head direction change angle, and a vehicle head direction change angular velocity; then the relationship between the driving behavior of the corresponding driving vehicle and the specific behavior is predicted based on the first driving vehicle data to obtain a plurality of second potential danger points, including:

[0018] The deep analysis model performs matching analysis on the deceleration data, the vehicle head direction change angle, and the vehicle head direction change angular velocity in the first driving vehicle data to determine the matching probability between the driving behavior of each corresponding driving vehicle and the specific behavior;

[0019] If the matching probability is higher than a threshold, the position of the corresponding moving vehicle is determined to be the second potential danger point.

[0020] Optionally, performing driving intervention on the vehicle at each of the first potential danger point and the second potential danger point includes:

[0021] Calculating the real-time distance between each of the first potential danger point, the second potential danger point and the vehicle;

[0022] If the real-time distance between the first potential danger point and the vehicle is less than a first distance threshold, a warning prompt is issued to the driver of the vehicle;

[0023] If the real-time distance between the second potential danger point and the vehicle is less than the second distance threshold, a braking or lane change avoidance instruction for the vehicle is generated and implemented, and the instruction is calculated taking into account ride comfort; if the real-time distance between the second potential danger point and the vehicle is greater than the second distance threshold, a warning prompt is issued to the driver of the vehicle.

[0024] The present invention also provides an image recognition system for assisted driving, the system comprising a vehicle-mounted camera and an analysis module, the vehicle-mounted camera being electrically connected to the analysis module;

[0025] The vehicle-mounted camera is used to obtain road structure image data within a first distance range ahead of the vehicle, and first running vehicle data within a second distance range ahead of the vehicle;

[0026] The analysis module is used to predict a number of first potential danger points based on the road structure image data; and predict the relationship between the driving behavior of the corresponding moving vehicle and the specific behavior based on the first moving vehicle data to obtain a number of second potential danger points, and perform driving intervention on the vehicle for each of the first potential danger points and the second potential danger points; wherein the first potential danger point and the second potential danger point are both related to undetected objects entering the road of the vehicle, and the specific behavior is the undetected object entering the road of the vehicle.

[0027] The present invention also provides an electronic device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program implements the method described in any of the preceding items when executed by the processor.

[0028] The present invention also provides a computer storage medium storing a computer program executable by a processor to implement any of the methods described above.

[0029] The present invention also provides a computer program product, which includes a computer program that can be executed by a processor to implement any of the methods described above.

[0030] The beneficial effects of the present invention are as follows: the present invention uses the on-board camera to analyze the road structure and the driving behavior of surrounding vehicles in advance, and takes appropriate driving intervention operations in time. On the one hand, it can effectively reduce the impact of "ghosting" on the driving safety of the vehicle; on the other hand, since the scheme adopted is to conduct predictive analysis in advance and perform driving intervention in time, it can also reduce the intensity of driving intervention, bringing a more comfortable driving experience to the passengers in the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 is a flow chart of an image recognition method for assisted driving disclosed in an embodiment of the present invention;

[0033] Figure 2 It is a structural schematic diagram of an image recognition system for assisted driving disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0035] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0036] like Figure 1 As shown, an embodiment of the present invention discloses an image recognition method for assisted driving, the method comprising the following steps:

[0037] Using a vehicle-mounted camera to obtain road structure image data within a first distance ahead of the vehicle, and predicting a number of first potential danger points based on the road structure image data;

[0038] Using a vehicle-mounted camera to obtain first traveling vehicle data within a second distance range ahead of the vehicle, predicting a relationship between a driving behavior of a corresponding traveling vehicle and a specific behavior based on the first traveling vehicle data, and obtaining a plurality of second potential danger points;

[0039] performing driving intervention on the vehicle at each of the first potential danger point and the second potential danger point;

[0040] Among them, the first potential danger point and the second potential danger point are both related to undetected objects that enter the road where the vehicle is traveling, and the specific behavior is that the undetected objects enter the road where the vehicle is traveling.

[0041] In order to solve the hazards caused by "ghost peeking", the present invention first uses the vehicle-mounted camera to perform image recognition on the image within the first distance range in front of the vehicle, and extracts the road structure image data. The road structure image data here refers to the physical structure of the road, from which the physical structure points where "ghost peeking" may occur can be determined, such as the gaps in the green belts on both sides of the lane, the zebra crossings at the intersections, and the special advance U-turn points for vehicles at the intersections. The above points are determined as the first potential danger points.

[0042] At the same time, the on-board camera is used to perform image recognition on images within the second distance range in front of the vehicle, and the moving vehicle data is extracted. Based on this, it is analyzed whether the driving behavior of the corresponding moving vehicle is related to a specific behavior (i.e. an undetected object intrudes into the vehicle's driving path, commonly known as "ghost poking its head"). If there is a correlation, the position of the corresponding vehicle is determined as the second potential danger point.

[0043] Therefore, the present invention uses the on-board camera to analyze the road structure and the driving behavior of surrounding vehicles in advance, and takes appropriate driving intervention operations in time. On the one hand, it can effectively reduce the impact of "ghost peeking" on the driving safety of the vehicle; on the other hand, since it adopts the solution of advance prediction and analysis and timely driving intervention, it can also reduce the intensity of driving intervention and bring a more comfortable driving experience to the passengers in the car.

[0044] Among them, the vehicle-mounted camera can be a binocular or monocular camera configured when the vehicle leaves the factory, or it can be the camera of a driving recorder.

[0045] Optionally, the method of using a vehicle-mounted camera to obtain road structure image data within a first distance range ahead of the vehicle includes:

[0046] Using a vehicle-mounted camera to obtain data of a second vehicle in a lane on the right side of the vehicle within a third distance range in front of the vehicle, wherein the second vehicle data includes the number of vehicles, the average speed of the vehicles, and the average distance between the vehicles;

[0047] The first distance range is determined according to the second driving vehicle data.

[0048] In this embodiment, when the vehicle is driving on internal roads such as non-residential areas, "ghost peeking" basically occurs from the right side of the vehicle, that is, pedestrians, non-motor vehicles, etc. will intrude from the non-visible area on the right side of the vehicle to the front of the vehicle. Considering the actual right-side driving road traffic rules, the present invention sets that during the vehicle driving on the highway (for example, the speed is greater than 10km / h), only the second driving vehicle data located in the right lane of the vehicle and in the third distance range in front is obtained. The obtained second driving vehicle data includes three types of data: the number of vehicles, the vehicle speed, and the average distance between vehicles. The above three types of data will affect the implementation of the intrusion behavior of the object located in the non-visible area on the right side of the vehicle. Based on this, the size of the first distance range can be determined. By adjusting the third distance range, the data processing load can be effectively reduced, and the prediction and response to the intrusion behavior will not be reduced.

[0049] Among them, the third distance range is a value smaller than the first distance range, for example, the first distance range can be 30m, 50m, and the third distance range can be 20m. Moreover, the third distance range can also be dynamically adjusted according to the type of road on which the vehicle is traveling. For example, the size of the third distance range is small when on a highway, medium when on a provincial road, and large when on a lower-level road. The reason for this setting is that the probability of pedestrians and non-motor vehicles breaking in is relatively low due to the closedness, semi-closedness, and high speed of high-level roads, while the speed of ordinary roads of low levels is generally very low, and the probability of pedestrians and non-motor vehicles breaking in is relatively high, and the size of the third distance range is adjusted accordingly.

[0050] Optionally, determining the first distance range according to the second driving vehicle data includes:

[0051] Substituting the number of vehicles, the average speed of the vehicles and the average distance between the vehicles into the following formula for calculation, the first distance range is obtained: ; In the formula, is the first distance range, is the number of vehicles, is the average speed of the vehicle, is the average distance between vehicles, , , is the corresponding dimensionless conversion function.

[0052] In this embodiment, the number of vehicles, the average speed of vehicles and the average distance between vehicles will affect the probability of pedestrians and non-motor vehicles to perform intrusion. Generally speaking, the more vehicles there are and the faster the average speed of vehicles, the more pedestrians, non-motor vehicles and other objects will significantly reduce their confidence and determination to perform intrusion in view of the significantly higher probability of collision risk, and the probability of performing intrusion will be correspondingly reduced, and there is no need to perform road structure analysis within a larger first distance range; and when the average distance between vehicles is larger, the confidence and determination of the above objects to perform intrusion will increase accordingly, because the risk of collision will increase appropriately, and it is necessary to perform road structure analysis within a larger first distance range accordingly. Therefore, based on the above rules, the present invention adopts the above calculation formula to comprehensively analyze the three data of the number of vehicles, the average speed of vehicles and the average distance between vehicles, so as to obtain the best first distance range.

[0053] Optionally, determining the first distance range according to the second driving vehicle data includes:

[0054] Obtaining a historical lane change probability of each lane in a driving area where the vehicle is located, and correcting the second driving vehicle data according to the historical lane change probability to obtain third driving vehicle data;

[0055] The first distance range is determined according to the third vehicle driving data.

[0056] In this embodiment, the second driving vehicle data obtained by the above detection is the real-time vehicle data that is already in the right lane of the vehicle, but there may be situations where the vehicle changes lanes from other lanes to the right lane of the vehicle and from the right lane of the vehicle to other lanes, which will also affect the real-time vehicle data in the right lane of the vehicle. Therefore, the present invention comprehensively analyzes the number of vehicles that change lanes from other lanes to the right lane of the vehicle and from the right lane of the vehicle to other lanes based on the historical lane change probability of each lane in the driving area where the vehicle is located, and accordingly corrects the second driving vehicle data obtained by the above detection, and uses the corrected third driving vehicle data to determine the first distance range.

[0057] The historical lane change probability is obtained based on the statistical analysis of the lane change history data of each lane in the area. The lane change history data can be obtained by the road monitoring system and vehicle network communication technology, and the details will not be elaborated here.

[0058] The above lane changing situation will directly affect the number of vehicles in the lane to the right of the vehicle. In addition, the lane changing behavior will also reduce the average speed of vehicles in the lane to the right of the vehicle to a certain extent. In addition, the lane changing behavior will also affect the average distance between vehicles. When there are more vehicles entering the lane than leaving the lane, the average distance between vehicles will become smaller. When there are fewer vehicles entering the lane than leaving the lane, the average distance between vehicles will become larger.

[0059] Therefore, based on the above historical lane change probability, the number of vehicles entering and exiting the lane on the right side of the vehicle is predicted, and new vehicle data is directly obtained based on this. Also, a correction coefficient is determined based on the total number of vehicles entering and exiting the lane and the corresponding negative proportional function, and the correction coefficient is multiplied by the original average vehicle speed to obtain a new average vehicle speed. Also, the size relationship between vehicles entering the lane change and vehicles exiting the lane change is calculated. When the number of vehicles entering the lane change is greater than the number of vehicles exiting the lane change, the correction coefficient of the average vehicle spacing is set to be less than 1, that is, the average vehicle spacing is lowered; when the number of vehicles exiting the lane change is greater than the number of vehicles entering the lane change, the correction coefficient of the average vehicle spacing is set to be greater than 1, that is, the average vehicle spacing is increased. Thus, the third driving vehicle data is obtained.

[0060] Optionally, the first driving vehicle data includes deceleration data, a vehicle head direction change angle, and a vehicle head direction change angular velocity; then the relationship between the driving behavior of the corresponding driving vehicle and the specific behavior is predicted based on the first driving vehicle data to obtain a plurality of second potential danger points, including:

[0061] The deep analysis model performs matching analysis on the deceleration data, the vehicle head direction change angle, and the vehicle head direction change angular velocity in the first driving vehicle data to determine the matching probability between the driving behavior of each corresponding driving vehicle and the specific behavior;

[0062] If the matching probability is higher than a threshold, the position of the corresponding moving vehicle is determined to be the second potential danger point.

[0063] In this embodiment, the intrusion of objects such as pedestrians and non-motor vehicles will cause the vehicle in front of the vehicle to decelerate suddenly. Based on this, it can be indirectly analyzed whether there is a "ghost peeking" behavior in front, and this "ghost peeking" behavior may also constitute a "ghost peeking" to the vehicle. In this regard, the present invention uses a deep analysis model to conduct a deep analysis of the driving behavior of the vehicle in front, so as to determine whether the driving behavior of the corresponding vehicle is caused by avoiding the intruding object. The driving behavior is mainly deceleration data, the angle of change of the direction of the head of the vehicle, and the angular velocity of the change of the direction of the head of the vehicle. The deceleration when avoiding the intruding object will be higher than the deceleration braking during normal driving. The angle of change of the direction of the head of the vehicle when avoiding the intruding object and the angular velocity of the change of the direction of the head of the vehicle will also be higher than the angle of change of the direction of the head of the vehicle during normal driving. The deep analysis model can be constructed based on deep learning algorithms such as SVM, CNN, LSTM, etc., and the present invention is not limited.

[0064] Optionally, performing driving intervention on the vehicle at each of the first potential danger point and the second potential danger point includes:

[0065] Calculating the real-time distance between each of the first potential danger point, the second potential danger point and the vehicle;

[0066] If the real-time distance between the first potential danger point and the vehicle is less than a first distance threshold, a warning prompt is issued to the driver of the vehicle;

[0067] If the real-time distance between the second potential danger point and the vehicle is less than the second distance threshold, a braking or lane change avoidance instruction for the vehicle is generated and implemented, and the instruction is calculated taking into account ride comfort; if the real-time distance between the second potential danger point and the vehicle is greater than the second distance threshold, a warning prompt is issued to the driver of the vehicle.

[0068] In this embodiment, the present invention responds to the first potential danger point and the second potential danger point separately. Specifically, the warning is output only when the real-time distance between the first potential danger point and the vehicle is less than the first distance threshold (for example, 15m), otherwise the warning remains silent. When the second potential danger point is far away from the vehicle (for example, more than 20m), the warning prompts the driver to pay attention, and when the second potential danger point is close to the vehicle (for example, more than 5m), the driver generally does not have time to react. At this time, the driving assistance system takes braking or lane change to avoid driving intervention, thereby ensuring driving safety. Moreover, the above instructions in the present invention are calculated based on the riding comfort of the occupants in the vehicle, and the riding comfort of the occupants in the vehicle is maintained to the greatest extent under the premise of ensuring effective response to the intrusion of pedestrians and non-motor vehicles.

[0069] like Figure 2As shown, the present invention also discloses an image recognition system for assisted driving, the system comprising a vehicle-mounted camera and an analysis module, the vehicle-mounted camera being electrically connected to the analysis module;

[0070] The vehicle-mounted camera is used to obtain road structure image data within a first distance range ahead of the vehicle, and first running vehicle data within a second distance range ahead of the vehicle;

[0071] The analysis module is used to predict a number of first potential danger points based on the road structure image data; and predict the relationship between the driving behavior of the corresponding moving vehicle and the specific behavior based on the first moving vehicle data to obtain a number of second potential danger points, and perform driving intervention on the vehicle for each of the first potential danger points and the second potential danger points; wherein the first potential danger point and the second potential danger point are both related to undetected objects entering the road of the vehicle, and the specific behavior is the undetected object entering the road of the vehicle.

[0072] The present invention also discloses an electronic device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program implements the method described in any of the preceding items when executed by the processor.

[0073] The present invention also discloses a computer storage medium, which stores a computer program that can be executed by a processor to implement any of the methods described in the preceding items.

[0074] The present invention also discloses a computer program product, which includes a computer program that can be executed by a processor to implement any of the methods described in the preceding items.

[0075] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0076] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0077] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or can be electrical, mechanical or other forms of connection.

[0078] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0079] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0080] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0081] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. An image recognition method for assisted driving, characterized in that: The method comprises the following steps: Using a vehicle-mounted camera to obtain road structure image data within a first distance ahead of the vehicle, and predicting a number of first potential danger points based on the road structure image data; Using a vehicle-mounted camera to obtain first traveling vehicle data within a second distance range ahead of the vehicle, predicting a relationship between a driving behavior of a corresponding traveling vehicle and a specific behavior based on the first traveling vehicle data, and obtaining a plurality of second potential danger points; performing driving intervention on the vehicle at each of the first potential danger point and the second potential danger point; Wherein, the first potential danger point and the second potential danger point are both related to an undetected object driving into the driving path of the vehicle, and the specific behavior is that the undetected object drives into the driving path of the vehicle; Use the vehicle-mounted camera to obtain road structure image data within the first distance range ahead of the vehicle, including: Using a vehicle-mounted camera to obtain data of a second vehicle in a lane on the right side of the vehicle within a third distance range in front of the vehicle, wherein the second vehicle data includes the number of vehicles, the average speed of the vehicles, and the average distance between the vehicles; Determine the first distance range according to the second driving vehicle data; Determining the first distance range according to the second driving vehicle data includes: Substituting the number of vehicles, the average speed of the vehicles and the average distance between the vehicles into the following formula for calculation, the first distance range is obtained: , where is the first distance range, is the number of vehicles, is the average speed of the vehicle, is the average distance between vehicles, , , is the corresponding dimensionless conversion function.

2. The image recognition method for assisted driving according to claim 1, characterized in that: Determining the first distance range according to the second driving vehicle data includes: Obtaining a historical lane change probability of each lane in a driving area where the vehicle is located, and correcting the second driving vehicle data according to the historical lane change probability to obtain third driving vehicle data; The first distance range is determined according to the third vehicle driving data.

3. The image recognition method for assisted driving according to claim 2, characterized in that: The first driving vehicle data includes deceleration data, a vehicle head direction change angle, and a vehicle head direction change angular velocity; then the relationship between the driving behavior of the corresponding driving vehicle and the specific behavior is predicted based on the first driving vehicle data to obtain a number of second potential danger points, including: The deep analysis model performs matching analysis on the deceleration data, the vehicle head direction change angle, and the vehicle head direction change angular velocity in the first driving vehicle data to determine the matching probability between the driving behavior of each corresponding driving vehicle and the specific behavior; If the matching probability is higher than a threshold, the position of the corresponding moving vehicle is determined to be the second potential danger point.

4. The image recognition method for assisted driving according to claim 1, characterized in that: Performing driving intervention on the vehicle for each of the first potential danger point and the second potential danger point includes: Calculating the real-time distance between each of the first potential danger point, the second potential danger point and the vehicle; If the real-time distance between the first potential danger point and the vehicle is less than a first distance threshold, a warning prompt is issued to the driver of the vehicle; If the real-time distance between the second potential danger point and the vehicle is less than the second distance threshold, a braking or lane change avoidance instruction for the vehicle is generated and implemented, and the instruction is calculated taking into account ride comfort; if the real-time distance between the second potential danger point and the vehicle is greater than the second distance threshold, a warning prompt is issued to the driver of the vehicle.

5. An image recognition system for assisted driving, the system being based on the method according to any one of claims 1 to 4, characterized in that: The system includes a vehicle-mounted camera and an analysis module, wherein the vehicle-mounted camera is electrically connected to the analysis module; The vehicle-mounted camera is used to obtain road structure image data within a first distance range ahead of the vehicle, and first running vehicle data within a second distance range ahead of the vehicle; The analysis module is used to predict a number of first potential danger points based on the road structure image data; and predict the relationship between the driving behavior of the corresponding moving vehicle and the specific behavior based on the first moving vehicle data to obtain a number of second potential danger points, and perform driving intervention on the vehicle for each of the first potential danger points and the second potential danger points; wherein the first potential danger point and the second potential danger point are both related to undetected objects entering the road of the vehicle, and the specific behavior is the undetected object entering the road of the vehicle.

6. An electronic device, characterized in that: The electronic device comprises: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the computer program implements the method according to any one of claims 1 to 4 when executed by the processor.

7. A computer storage medium, characterized in that: The computer storage medium stores a computer program that can be executed by a processor to implement the method according to any one of claims 1 to 4.

8. A computer program product, characterized in that: The computer program product comprises a computer program executable by a processor to implement the method according to any one of claims 1 to 4.