Vehicle blind area early warning method and device, electronic equipment and storage medium

Abstract

The invention discloses a vehicle blind area early warning method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the position prediction processing of a current obstacle object based on the first fusion driving data of the fusion driving data of the current obstacle object at each historical moment in a historical time period; the fused driving data at each historical moment is based on a first weight and a second weight; the data fusion module is obtained by fusing first filtered data obtained by performing abnormal filtering on first driving data extracted based on an image acquired by a camera device and second filtered data obtained by performing abnormal filtering on second driving data acquired based on a preset sensor; the predicted position information, the regional position information and the current environment data are input into a blind area early warning recognition model for blind area early warning recognition; and performing blind area early warning processing based on a blind area early warning identification result. According to the technical scheme provided by the invention, the fusion of the driving data from different sources can be performed in a targeted manner, and the driving safety is improved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a vehicle blind spot warning method, device, electronic device and storage medium. Background Technology

[0002] With the rapid development of the automotive industry and the continuous improvement of vehicle performance, vehicle safety is receiving increasing attention. Blind spots can occur during driving, and drivers are prone to accidents due to these blind spots. Related technologies detect obstacles such as pedestrians and vehicles in blind spots and use the obstacle's driving data to provide blind spot warnings. However, the driving data collected by these technologies is often inaccurate, making it difficult to accurately predict the location of obstacles, which in turn leads to poor accuracy in blind spot warnings and reduced driving safety. Summary of the Invention

[0003] This application provides a vehicle blind spot warning method, device, equipment, storage medium, and computer program product, which can selectively fuse driving data from different sources to improve the effectiveness and accuracy of the fused driving data, thereby effectively improving the accuracy of blind spot warning and enhancing driving safety.

[0004] On the one hand, this application provides a vehicle blind spot warning method, the method comprising:

[0005] When a blind spot occurs during the current vehicle's driving, the system acquires first fused driving data, current environmental data, and the location information of the blind spot corresponding to the current obstacle object within a preset area corresponding to the current vehicle. The first fused driving data includes fused driving data of the current obstacle object at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing first filtered data and second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering the first driving data of the current obstacle object at each historical moment for abnormal driving data, and the second filtered data is obtained by filtering the second driving data of the current obstacle object at each historical moment for abnormal driving data. The first weight and the second weight are determined based on the number of occurrences of abnormal driving data of the current obstacle object within a first preset time period before each historical moment. The first driving data is driving data extracted from images collected by a camera device, and the second driving data is driving data collected by a preset sensor.

[0006] Based on the first fused driving data, the current obstacle object is subjected to position prediction processing to obtain the predicted position information of the current obstacle object after a preset time after the current moment;

[0007] The predicted location information, the regional location information, and the current environmental data are input into the blind spot early warning identification model to perform blind spot early warning identification and obtain the blind spot early warning identification result.

[0008] Based on the blind spot warning identification result, perform blind spot warning processing for the current vehicle.

[0009] On the other hand, a vehicle blind spot warning device is provided, the device comprising:

[0010] The first data acquisition module is configured to, when a blind spot occurs during the current vehicle's driving, acquire first fused driving data, current environmental data, and the area location information corresponding to the blind spot within a preset area corresponding to the current vehicle. The first fused driving data includes fused driving data of the current obstacle at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing first filtered data and second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering the first driving data of the current obstacle at each historical moment for abnormal driving data, and the second filtered data is obtained by filtering the second driving data of the current obstacle at each historical moment for abnormal driving data. The first weight and the second weight are determined based on the number of occurrences of abnormal driving data of the current obstacle within a first preset time period before each historical moment. The first driving data is driving data extracted from images collected by a camera device, and the second driving data is driving data collected by a preset sensor.

[0011] The location prediction module is configured to perform location prediction processing on the current obstacle object based on the first fused driving data, and obtain the predicted location information of the current obstacle object after a preset time after the current time.

[0012] The blind spot warning and identification module is configured to input the predicted location information, the area location information and the current environmental data into the blind spot warning and identification model to perform blind spot warning and identification, and obtain the blind spot warning and identification result.

[0013] The blind spot warning processing module is configured to perform blind spot warning processing on the current vehicle based on the blind spot warning recognition result.

[0014] On the other hand, an electronic device is provided, including: a processor;

[0015] Memory used to store the processor's executable instructions;

[0016] The processor is configured to execute the instructions to implement the vehicle blind spot warning method described in any of the preceding claims.

[0017] On the other hand, a computer-readable storage medium is provided, which, when the instructions in the storage medium are executed by the processor of an electronic device, enables the electronic device to perform any of the above-described vehicle blind spot warning methods.

[0018] On the other hand, a computer program product or computer program is provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the vehicle blind spot warning method provided in the various optional implementations described above.

[0019] The vehicle blind spot warning method, device, equipment, storage medium, and computer program product provided in this application have the following technical effects:

[0020] This application, when a blind spot occurs during the current vehicle's operation, acquires first fused driving data, current environmental data, and the location information of the blind spot corresponding to the obstacle object within a preset area corresponding to the current vehicle. The first fused driving data includes fused driving data of the obstacle object at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing first filtered data and second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering abnormal driving data from the first driving data extracted from images of the obstacle object at each historical moment based on images captured by a camera device. The second filtered data is obtained by filtering abnormal driving data from the second driving data of the obstacle object at each historical moment based on images captured by a preset sensor. This can be achieved by filtering abnormal driving data from different sources (camera device and preset sensor). The system integrates driving data from various sources to effectively improve the accuracy of driving data. The first and second weights are determined based on the number of times abnormal driving data of the current obstacle object appears in a first preset time period before each historical moment. This allows for more targeted integration of driving data from different sources, thereby improving the effectiveness and accuracy of the integrated driving data. Based on the first integrated driving data, the system performs position prediction processing on the current obstacle object to obtain the predicted position information of the current obstacle object at a preset time after the current moment. This allows for accurate prediction of the obstacle object's position. The predicted position information, regional position information, and current environmental data are then input into the blind spot warning recognition model for blind spot warning recognition. Based on the blind spot warning recognition results, the system performs blind spot warning processing on the current vehicle, which effectively improves the accuracy of blind spot warnings and thus enhances driving safety. Attached Figure Description

[0021] To more clearly illustrate the technical solutions and advantages in the embodiments of 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 only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the application environment of a vehicle blind spot warning method provided in an embodiment of this application;

[0023] Figure 2 This is a flowchart illustrating a vehicle blind spot warning method provided in an embodiment of this application;

[0024] Figure 3 This is a schematic diagram of a process for fusing first and second driving data of obstacles within a preset area of ​​the current vehicle at each moment, as provided in an embodiment of this application.

[0025] Figure 4 This is a schematic diagram of the structure of a vehicle blind spot warning device provided in an embodiment of this application;

[0026] Figure 5 This is a block diagram of an electronic device for vehicle blind spot warning provided in an embodiment of this application;

[0027] Figure 6 This is a block diagram of another electronic device for vehicle blind spot warning provided in an embodiment of this application. Detailed Implementation

[0028] The technical solutions of the embodiments 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, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0030] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0031] Please see Figure 1 , Figure 1 This is a schematic diagram of the application environment of a vehicle blind spot warning method provided in an embodiment of this application. The application environment may include at least an in-vehicle terminal 100 and a server 200.

[0032] In an optional embodiment, the vehicle terminal 100 can be used to collect driving data corresponding to the vehicle and obstacles near the vehicle, and send the driving data to the server 200. The server 200 can store the driving data, combine the driving data to perform blind spot warnings for the vehicle, and feed back the corresponding warning information to the vehicle terminal 100.

[0033] In one specific embodiment, server 200 can be an independent physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.

[0034] In addition, it should be noted that, Figure 1 The example shown is merely an application environment for a vehicle blind spot warning method. The embodiments in this specification are not limited to the above. For example, the storage of driving data and the use of driving data to perform vehicle blind spot warning can also be implemented by edge computing devices.

[0035] In the embodiments described in this specification, the vehicle terminal 100 and the server 200 can be directly or indirectly connected via wired or wireless communication, and this application does not impose any restrictions on this.

[0036] The following describes a vehicle blind spot warning method according to this application. Figure 2 This is a flowchart illustrating a vehicle blind spot warning method provided in an embodiment of this application. This specification provides the method operation steps as shown in the embodiments or flowcharts, but based on conventional or non-inventive labor, more or fewer operation steps may be included. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual system or server product execution, the method can be executed sequentially according to the embodiments or drawings, or in parallel (e.g., in a parallel processor or multi-threaded processing environment). Specifically, as shown... Figure 2 As shown, the method may include:

[0037] S201: When a blind spot occurs during the current vehicle's driving, obtain the first fused driving data, current environmental data, and the area location information corresponding to the blind spot within the preset area corresponding to the current vehicle.

[0038] In one specific embodiment, a blind spot can be an area that the driver cannot observe directly or through the in-vehicle mirror system during the current vehicle driving process. It is usually a blind spot caused by the vehicle's structural design, or an area that cannot be effectively monitored due to the failure of the vehicle's camera equipment or sensors. The aforementioned first fused driving data includes fused driving data of the current obstacle object at each historical moment within a historical time period. Specifically, the current obstacle object may include at least one obstacle object within a preset area, such as a pedestrian or vehicle. The fused driving data at each historical moment within the aforementioned historical time period is obtained by fusing the first filtered data and the second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering the first driving data of the current obstacle object at each historical moment for abnormal driving data, and the second filtered data is obtained by filtering the second driving data of the current obstacle object at each historical moment for abnormal driving data. The first weight and the second weight are determined based on the number of occurrences of abnormal driving data of the current obstacle object within a first preset time period prior to each historical moment. The first driving data is driving data extracted from images captured by a camera device, and the second driving data is driving data collected by a preset sensor (non-visual sensor). Specifically, the driving data may include speed, acceleration, and geographical location information. Accordingly, the fused driving data includes: speed fused data, acceleration fused data, and location fused data.

[0039] In one specific embodiment, the preset area can be a designated area centered on the current vehicle (i.e., the surrounding area), which can be specifically set according to the actual application. Optionally, the preset sensor can be an ultrasonic sensor, which can be used to detect the distance and orientation of obstacles within the preset area. By emitting multiple ultrasonic pulses and measuring the time difference of the reflected pulses, the relative distance change of the obstacles can be detected. Therefore, given its own geographical location information (obtained by combining GPS positioning), the geographical location information, speed, and acceleration of the surrounding obstacles can be calculated based on the detection results of the ultrasonic sensor. Accordingly, the first driving data of the current obstacle at each moment can be obtained.

[0040] Furthermore, it should be noted that combining ultrasonic sensors to collect the first driving data corresponding to the current obstacle is just an example. In practical applications, other sensors can also be used to collect only the first driving data.

[0041] In one specific embodiment, during the current vehicle's movement, images (videos) can be captured from a preset area. A pre-trained target detection model can be used to detect targets in the images captured by the camera, determining the position and size of the obstacle in the image. Furthermore, combining the obstacle's position and size with the camera's geographical location and device information, the geographical location of the obstacle is determined. The target detection model can be an artificial intelligence model for quickly detecting and identifying targets (obstacles) in images. Further, multiple images containing the obstacle can be combined to perform trajectory analysis on the obstacle, determining its speed and acceleration. Accordingly, second driving data for the obstacle at each moment can be obtained.

[0042] In an optional embodiment, during the current vehicle's driving process, the first and second driving data of obstacles within a preset area of ​​the current vehicle can be fused in real time at each moment; correspondingly, such as Figure 3 As shown, the above method may further include:

[0043] S301: Obtain the first driving data and the second driving data of the current obstacle object at each historical moment;

[0044] S303: Perform anomaly detection on the first driving data and the second driving data respectively, and determine the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data;

[0045] S305: Based on the first anomaly detection result and the second anomaly detection result, perform anomaly data filtering processing on the first driving data and the second driving data respectively to obtain the first filtered data and the second filtered data.

[0046] S307: Based on the first and second anomaly detection results of the current obstacle object within a first preset time period before each historical moment, determine the first occurrence count and the second occurrence count respectively;

[0047] S309: Determine the first weight and the second weight based on the first occurrence count and the second occurrence count;

[0048] S311: Based on the first weight and the second weight, the first filtered data and the second filtered data are fused to obtain the fused driving data of the current obstacle object at each historical moment.

[0049] In an optional embodiment, the above-mentioned anomaly detection of the first driving data and the second driving data, and the determination of the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data, may include:

[0050] Get the first mean driving data of the current obstacle object within a second preset time period before each historical moment, the first standard deviation of the first driving data within the second preset time period, the second mean driving data within the second preset time period, and the second standard deviation of the second driving data within the second preset time period.

[0051] Determine the first driving difference between the first driving data and the first driving mean of the current obstacle object at each historical moment;

[0052] Determine the second driving difference between the second driving data and the second driving mean of the current obstacle object at each historical moment;

[0053] The first anomaly detection result is determined based on the first driving difference and the first standard deviation;

[0054] The second anomaly detection result is determined based on the second driving difference and the second standard deviation.

[0055] In one specific embodiment, the second preset time period can be preset according to actual application. The number of first and second driving data collected within the second preset time period can be determined by combining the collection frequency and the duration of the second preset time period. For example, the second preset time period is the first 10 seconds of each historical moment, and the collection frequency is 1 time / 1 second. Accordingly, the number of first and second driving data collected in the 10 seconds before each historical moment is 10 each. The first driving mean corresponding to each historical moment can be the mean of multiple first driving data collected within the second preset time period before that historical moment; the second driving mean corresponding to each historical moment can be the mean of multiple second driving data collected within the second preset time period before that historical moment; the first standard deviation corresponding to each historical moment can be the standard deviation of multiple first driving data collected within the second preset time period before that historical moment; and the second standard deviation corresponding to each historical moment can be the standard deviation of multiple second driving data collected within the second preset time period before that historical moment. Specifically, the driving data includes speed, acceleration, and geographical location information. Correspondingly, the first driving mean, first standard deviation, second driving mean, and second standard deviation mentioned above can be the first driving mean, first standard deviation, second driving mean, and second standard deviation corresponding to speed, acceleration, and geographical location information, respectively.

[0056] In one specific embodiment, the first driving difference value corresponding to each historical moment may include the speed difference, position difference, and acceleration difference between the first driving data of the current obstacle object and the first driving average value at each historical moment. The first second driving difference value corresponding to each historical moment may include the speed difference, position difference, and acceleration difference between the second driving data of the current obstacle object and the second driving average value at each historical moment.

[0057] In a specific embodiment, determining the first anomaly detection result based on the first driving difference and the first standard deviation may include, for any one of the first driving differences (speed difference, position difference, or acceleration difference), if the difference is greater than a preset multiple of the corresponding first standard deviation (speed standard deviation, position standard deviation, or acceleration standard deviation), then the corresponding driving data (speed, geographical location information, or acceleration) in the first driving data corresponding to that difference is considered abnormal driving data; conversely, if the difference is less than that, the corresponding driving data (speed, geographical location information, or acceleration) in the first driving data corresponding to that difference is considered normal driving data (non-abnormal driving data). Accordingly, the aforementioned first anomaly detection result can be used to indicate whether each driving data item (speed, geographical location information, or acceleration) in the first driving data corresponding to each historical moment is abnormal driving data. Specifically, the preset multiple can be set according to actual application, for example, 3 times.

[0058] In one specific embodiment, the detailed process of determining the second anomaly detection result based on the second driving difference and the second standard deviation can be found in the detailed process of determining the first anomaly detection result based on the first driving difference and the first standard deviation in one specific embodiment, which will not be repeated here.

[0059] In the above embodiments, by combining the mean of the driving data of the current obstacle object in the second preset time period before each historical moment with the driving difference between the driving data of the current obstacle object in each historical moment and the first standard deviation of the driving data in the second preset time period, abnormal driving data can be effectively identified by combining the deviation of the driving data of each historical moment from the mean of a certain time period with the standard deviation of that time period, thereby ensuring the effectiveness of the driving data for subsequent location prediction.

[0060] In another optional embodiment, the above-mentioned anomaly detection of the first driving data and the second driving data, and the determination of the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data, may include:

[0061] Obtain the first driving data and the second driving data of the current obstacle object within a third preset time period at each historical moment;

[0062] The first driving data within the third preset time period and the second driving data within the third preset time period are input into the preset anomaly recognition model for anomaly recognition, and the first anomaly detection result and the second anomaly detection result are obtained.

[0063] In a specific embodiment, each historical moment corresponds to a third preset time period including each historical moment, and the preset anomaly recognition model can be an artificial intelligence model used for abnormal driving data recognition; specifically, each piece of driving data corresponds to a preset anomaly recognition model.

[0064] In the above embodiments, by combining a preset anomaly detection model to detect anomalies in driving data over a period of time, the model can learn the distribution of normal driving data, thereby detecting abnormal driving data that deviates from the normal distribution, ensuring the effectiveness of driving data for subsequent location prediction.

[0065] In a specific embodiment, the above-mentioned abnormal data filtering processing of the first driving data and the second driving data based on the first anomaly detection result and the second anomaly detection result to obtain the first filtered data and the second filtered data may include: filtering out abnormal driving data in the first driving data according to the first anomaly detection result to obtain the first filtered data corresponding to each historical time; and filtering out abnormal driving data in the second driving data according to the second anomaly detection result to obtain the second filtered data corresponding to each historical time.

[0066] In a specific embodiment, determining the first occurrence count and the second occurrence count based on the first and second anomaly detection results of the current obstacle object within a first preset time period before each historical moment may include:

[0067] Based on the first anomaly detection result of the current obstacle object within the first preset time period before each historical moment, determine the first occurrence count;

[0068] The second occurrence count is determined based on the second anomaly detection results of the current obstacle object within a first preset time period prior to each historical moment.

[0069] In a specific embodiment, the first occurrence count can be the number of times abnormal driving data of the current obstacle object appears in the first driving data within a first preset time period before each historical moment; the second occurrence count can be the number of times abnormal driving data of the current obstacle object appears in the second driving data within a first preset time period before each historical moment.

[0070] In one specific embodiment, the first weight represents the effectiveness of the first filtered data, and the second weight represents the effectiveness of the second filtered data; the first weight is negatively correlated with the first occurrence coefficient and positively correlated with the second occurrence frequency; the second weight is positively correlated with the first occurrence coefficient and negatively correlated with the second occurrence frequency.

[0071] In a specific embodiment, when the first occurrence count is not zero, the determination of the first weight and the second weight based on the first occurrence count and the second occurrence count can be combined with the following formula.

[0072] α t1 =1-α t2 ,

[0073] Where, α t1 α represents the first weight at time t. t2 k represents the second weight at time t. t1 k represents the number of times the first occurrence occurs in the first preset time period before time t. t2This indicates the number of times the second occurrence occurs in the first preset time period before time t.

[0074] If the first occurrence count is zero, and the second occurrence count is also zero, then both the first and second weights are 0.5. If the first occurrence count is zero, and the second occurrence count is not zero, then the values ​​of the first and second weights can be preset, and the sum of the first and second weights is equal to 1, but the first weight is greater than the second weight.

[0075] In a specific embodiment, fusing the first filtered data and the second filtered data based on the first weight and the second weight to obtain the fused driving data of the current obstacle object at each historical moment may include: weighting and summing the first filtered data and the second filtered data based on the first weight and the second weight to obtain the fused driving data of the current obstacle object at each historical moment.

[0076] Furthermore, it should be noted that if either the first or second driving data at a certain time t is completely missing, the weight corresponding to the partially missing driving data is set to 1, and the weight corresponding to the completely missing driving data is set to 0. If both the first and second driving data at a certain time t are completely missing, the driving data at that time t is empty and no fusion is performed.

[0077] In the above embodiments, anomaly detection is performed on the first driving data extracted from images captured by camera equipment and the second driving data collected by sensors at each historical moment for the current obstacle object. The anomaly detection results are then combined to filter the abnormal driving data, ensuring the validity of the filtered driving data. The anomaly detection results are also used to determine the first occurrence count of abnormal driving data in the first driving data extracted from images captured by camera equipment and the second occurrence count of abnormal driving data in the second driving data collected by sensors within a certain period prior to each historical time. The first and second occurrence counts are then combined to determine the corresponding weights of the filtered driving data from different sources (camera equipment and non-visual sensors), effectively ensuring the validity and accuracy of the fused driving data obtained by fusing the filtered driving data from different sources, thereby improving the accuracy of subsequent location prediction.

[0078] S203: Based on the first fused driving data, perform position prediction processing on the current obstacle object to obtain the predicted position information of the current obstacle object at a preset time after the current moment.

[0079] In one specific embodiment, the preset time can be set according to the actual application, such as 3 seconds after the current moment; in an optional embodiment, the above-mentioned position prediction processing of the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object within a second preset time period after the current moment may include:

[0080] Based on the first fused driving data, determine the motion pattern corresponding to the current obstacle object;

[0081] Based on the motion pattern and the first fused driving data, the current obstacle object is subjected to position prediction processing to obtain the predicted position information.

[0082] In a specific embodiment, the motion mode corresponding to the current obstacle object can be a uniform linear motion mode, an acceleration motion mode (motion mode of acceleration or deceleration) or other motion modes (motion models other than uniform linear motion mode and acceleration motion mode).

[0083] In a specific embodiment, when the motion mode corresponding to the current obstacle object is a uniform linear motion mode or an acceleration motion mode, the position prediction processing can be performed by combining the first fused driving data of the current obstacle object (each obstacle object) and using a uniform linear motion model or an acceleration motion model; when the motion mode corresponding to the current obstacle object is another motion mode, the position prediction processing can be performed by combining the first fused driving data and using a Bayesian method.

[0084] In the above embodiments, during the location prediction process, the motion pattern corresponding to the current obstacle object is first determined by combining the fused driving data, and then the location prediction processing of the current obstacle object is performed by combining the motion pattern and the first fused driving data. This allows for targeted location prediction and improves the accuracy of location prediction.

[0085] S205: Input the predicted location information, regional location information and current environmental data into the blind spot early warning identification model to perform blind spot early warning identification and obtain the blind spot early warning identification result.

[0086] In one specific embodiment, the blind spot warning recognition model can be a pre-trained artificial intelligence model for blind spot warning recognition. Specifically, the blind spot warning recognition result can be used to indicate the probability that the current obstacle object will enter the blind spot at a preset time. Specifically, when the current obstacle object includes multiple obstacle objects, the blind spot warning recognition result can be used to indicate the probability that each of the multiple obstacle objects will enter the blind spot at a preset time.

[0087] S207: Based on the blind spot warning recognition results, perform blind spot warning processing for the current vehicle.

[0088] In an optional embodiment, the above-mentioned blind spot warning processing for the current vehicle based on the blind spot warning recognition result includes:

[0089] If the blind spot warning identification result indicates that the probability of the current obstacle entering the blind spot within a preset time is greater than or equal to the first preset threshold, the second preset warning information is fed back.

[0090] In one specific embodiment, the second preset warning information can be information that prompts the driver that an obstacle will enter the blind spot. Specifically, it includes the predicted location information of the obstacle. Specifically, the warning form of the second preset warning information can include: sound and light warning (reminding the driver through sound or visual signals inside the vehicle), and tactile warning (adding a tactile feedback device to the driver's seat or steering wheel to remind the driver of potential blind spot risks through vibration).

[0091] In addition, if the probability of an obstacle entering the blind spot within a preset time is greater than or equal to the third preset threshold, the braking system can be automatically triggered to slow down or stop the vehicle to avoid a collision. Specifically, the third preset threshold is greater than the first preset threshold. The first and third preset thresholds can be set according to actual application requirements.

[0092] In a specific embodiment, when the current obstacle object includes multiple obstacle objects, the above-mentioned feedback of the second preset warning information when the blind spot warning identification result indicates that the probability of the current obstacle object entering the blind spot within a preset time is greater than or equal to the first preset threshold may include: when the blind spot warning identification result indicates that the probability of any obstacle object entering the blind spot within a preset time is greater than or equal to the first preset threshold, feedback of the second preset warning information corresponding to any obstacle object.

[0093] If the blind spot warning identification result indicates that the probability of the current obstacle entering the blind spot within a preset time is less than the first preset threshold, no warning may be issued;

[0094] In an optional embodiment, the above-mentioned blind spot warning processing for the current vehicle based on the blind spot warning recognition result may include:

[0095] If the blind spot warning and identification results indicate that the probability of the current obstacle entering the blind spot within a preset time is less than a first preset threshold, the current fused driving data of the current vehicle at the current moment is obtained;

[0096] The current fused driving data and current environmental data are input into the vehicle accident recognition model to identify vehicle accidents and obtain vehicle accident recognition results.

[0097] If the vehicle accident identification result indicates that the probability of the current vehicle being involved in an accident is greater than or equal to the second preset threshold, the first preset warning information will be fed back.

[0098] In one specific embodiment, during the current driving process, GPS, speed sensors, and acceleration sensors can be used to collect the vehicle's geographical location, speed, and acceleration at each moment. Alternatively, images of the vehicle can be captured by a camera, and the vehicle's geographical location, speed, and acceleration at each moment can be extracted from the images. This allows for the fusion of driving data collected by sensors and driving data extracted from images captured by cameras. For details on the process of acquiring the current fused driving data of the vehicle at the current moment, please refer to the relevant descriptions above, which will not be repeated here.

[0099] In one specific embodiment, the vehicle accident recognition model can be a pre-trained artificial intelligence model for vehicle accident recognition. Specifically, the vehicle accident recognition result can be used to indicate the probability that the current vehicle has been involved in an accident. The second preset threshold can be set according to the actual application.

[0100] In a specific embodiment, the details of the first preset warning information can be found in the relevant description of the second preset warning information above, and will not be repeated here.

[0101] In the above embodiments, when the blind spot warning recognition result indicates that the probability of the current obstacle entering the blind spot within a preset time is greater than or equal to a first preset threshold, a second preset warning message is fed back; when the blind spot warning recognition result indicates that the probability of the current obstacle entering the blind spot within a preset time is less than the first preset threshold, the current fused driving data of the current vehicle at the current moment is obtained, and the current fused driving data and the current environmental data are input into the vehicle accident recognition model to perform vehicle accident recognition, thereby obtaining the vehicle accident recognition result. When the vehicle accident recognition result indicates that the probability of the current vehicle causing an accident is greater than or equal to the second preset threshold, the first preset warning message is fed back. The blind spot warning recognition result and the vehicle accident recognition result can be combined to perform dual blind spot warning processing, thereby improving driving safety.

[0102] In an optional embodiment, the above method may further include:

[0103] Perform anomaly detection on the first fused driving data to determine the third anomaly detection result;

[0104] Accordingly, the above-mentioned location prediction processing of the current obstacle object based on the first fused driving data, to obtain the predicted location information of the current obstacle object within a second preset time period after the current moment, includes:

[0105] If the third anomaly detection result indicates that there is no abnormal driving data in the first fused driving data, the step of performing position prediction processing on the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object in the second preset time period after the current time is executed.

[0106] In a specific embodiment, the detailed process of performing anomaly detection on the first fused driving data and determining the third anomaly detection result can be found in the above-mentioned detailed process of performing anomaly detection on the first driving data and the second driving data respectively and determining the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data. It will not be repeated here.

[0107] In the above embodiments, after obtaining the first fused driving data obtained by fusing driving data collected by sensors and driving data extracted from images collected by camera devices, anomaly detection is performed on the first fused driving data. If the anomaly detection result indicates that there is no abnormal driving data in the first fused driving data, position prediction processing is performed on the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object within a second preset time period after the current moment. This can better ensure the reliability of the driving data for position prediction, thereby improving the accuracy of position prediction and enhancing driving safety.

[0108] In an optional embodiment, the above method may further include:

[0109] If the third anomaly detection result indicates that there is abnormal driving data in the first fused driving data, the first fused driving data is filtered for abnormal data based on the third anomaly detection result to obtain the second fused driving data.

[0110] Based on the interpolation method, the second fused driving data is corrected to obtain the third fused driving data;

[0111] Accordingly, the above-mentioned location prediction processing of the current obstacle object based on the first fused driving data to obtain the predicted location information of the current obstacle object within a second preset time period after the current moment may include:

[0112] Based on the third-party fusion driving data, the current obstacle object is subjected to position prediction processing to obtain the predicted position information.

[0113] In one specific embodiment, the process of filtering abnormal data from the first fused driving data based on the third anomaly detection result to obtain the second fused driving data may include: filtering out abnormal driving data from the first fused driving data based on the third anomaly detection result to obtain the second fused driving data.

[0114] In a specific embodiment, the process of correcting the second fused driving data based on interpolation to obtain the third fused driving data may include: if the driving data at a certain moment in the second fused driving data is empty, the average value of the driving data at that moment and the driving data at subsequent moments may be used as the driving data at that moment, thus obtaining the aforementioned third fused driving data.

[0115] In a specific embodiment, the location prediction processing of the current obstacle object is performed based on the third fused driving data. The specific details of the predicted location information can be found in the above-mentioned details of the location prediction processing of the current obstacle object combined with the second fused driving data, which will not be repeated here.

[0116] In the above embodiments, when the third anomaly detection result indicates that there is abnormal driving data in the first fused driving data, the first fused driving data is filtered based on the third anomaly detection result to obtain the second fused driving data, thereby realizing secondary abnormal driving data monitoring and filtering. Based on the interpolation method, the second fused driving data is corrected to obtain the third fused driving data. This can better improve the effectiveness and accuracy of the fused driving data, thereby improving the accuracy of the location prediction processing of the current obstacle object based on the fused driving data, and better improving the accuracy of the warning and driving safety.

[0117] As can be seen from the technical solutions provided in the embodiments of this specification above, when a blind spot occurs during the current vehicle's driving process, this specification obtains the first fused driving data, current environmental data, and the area location information corresponding to the blind spot within a preset area corresponding to the current vehicle. The first fused driving data includes the fused driving data of the current obstacle at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing the first filtered data and the second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering abnormal driving data from the first driving data extracted from the image collected by the camera device at each historical moment of the current obstacle. The second filtered data is obtained by filtering abnormal driving data from the second driving data collected by a preset sensor at each historical moment of the current obstacle. This can be achieved by filtering abnormal driving data from different sources. The driving data (from camera equipment and preset sensors) is fused to effectively improve the accuracy of the driving data. The first and second weights are determined based on the number of times abnormal driving data of the current obstacle object appears in a first preset time period before each historical moment. This allows for more targeted fusion of driving data from different sources, thereby improving the effectiveness and accuracy of the fused driving data. Based on the first fused driving data, the location of the current obstacle object is predicted to obtain the predicted location information of the current obstacle object at a preset time after the current moment. This allows for accurate prediction of the obstacle object's location. The predicted location information, regional location information, and current environmental data are then input into the blind spot warning recognition model for blind spot warning recognition. Based on the blind spot warning recognition results, the current vehicle's blind spot warning is processed, which effectively improves the accuracy of blind spot warning and thus enhances driving safety.

[0118] This application also provides a vehicle blind spot warning device, such as... Figure 4 As shown, the above-mentioned device includes:

[0119] The first data acquisition module 410 is configured to acquire, when a blind spot occurs during the current vehicle's driving, first fused driving data, current environmental data, and the area location information corresponding to the blind spot within a preset area corresponding to the current vehicle. The first fused driving data includes fused driving data of the current obstacle at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing first filtered data and second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering the first driving data of the current obstacle at each historical moment for abnormal driving data, and the second filtered data is obtained by filtering the second driving data of the current obstacle at each historical moment for abnormal driving data. The first weight and the second weight are determined based on the number of occurrences of abnormal driving data of the current obstacle within a first preset time period before each historical moment. The first driving data is driving data extracted from images collected by a camera device, and the second driving data is driving data collected by a preset sensor.

[0120] The position prediction module 420 is configured to perform position prediction processing on the current obstacle object based on the first fused driving data, and obtain the predicted position information of the current obstacle object after a preset time after the current time.

[0121] The blind spot warning and identification module 430 is configured to input the predicted location information, the area location information and the current environmental data into the blind spot warning and identification model to perform blind spot warning and identification, and obtain the blind spot warning and identification result.

[0122] The blind spot warning processing module 440 is configured to perform blind spot warning processing on the current vehicle based on the blind spot warning recognition result.

[0123] In an optional embodiment, the apparatus further includes:

[0124] The first anomaly detection module is configured to perform anomaly detection on the first fused driving data and determine the third anomaly detection result.

[0125] The location prediction module 420 is further configured to perform the step of performing location prediction processing on the current obstacle object based on the first fused driving data, and obtaining the predicted location information of the current obstacle object within a second preset time period after the current time, when the third anomaly detection result indicates that there is no abnormal driving data in the first fused driving data.

[0126] In an optional embodiment, the apparatus further includes:

[0127] The first anomaly filtering module is configured to perform anomaly filtering processing on the first fused driving data based on the third anomaly detection result when the third anomaly detection result indicates that there is abnormal driving data in the first fused driving data, so as to obtain the second fused driving data.

[0128] The correction processing module is configured to perform correction processing on the second fused driving data based on interpolation to obtain the third fused driving data.

[0129] The location prediction module 420 is further configured to perform location prediction processing on the current obstacle object based on the third fused driving data to obtain the predicted location information.

[0130] In an optional embodiment, the apparatus further includes:

[0131] The second data acquisition module is configured to acquire the first driving data of the current obstacle object at each historical moment and the second driving data of the current obstacle object at each historical moment.

[0132] The second anomaly detection module is configured to perform anomaly detection on the first driving data and the second driving data respectively, and determine the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data.

[0133] The second anomaly filtering module is configured to perform anomaly data filtering processing on the first driving data and the second driving data based on the first anomaly detection result and the second anomaly detection result, respectively, to obtain the first filtered data and the second filtered data.

[0134] The occurrence count determination module is configured to execute a first occurrence count and a second occurrence count based on the first anomaly detection result and the second anomaly detection result of the current obstacle object within the first preset time period before each historical moment. The first occurrence count is the number of times abnormal driving data of the current obstacle object appears in the first driving data within the first preset time period before each historical moment; the second occurrence count is the number of times abnormal driving data of the current obstacle object appears in the second driving data within the first preset time period before each historical moment.

[0135] The weight determination module is configured to determine a first weight and a second weight based on the first occurrence count and the second occurrence count. The first weight represents the effectiveness of the first filtered data, and the second weight represents the effectiveness of the second filtered data. The first weight is negatively correlated with the first occurrence coefficient and positively correlated with the second occurrence count; the second weight is positively correlated with the first occurrence coefficient and negatively correlated with the second occurrence count.

[0136] The fusion processing module is configured to perform fusion processing on the first filtered data and the second filtered data based on the first weight and the second weight, to obtain the fused driving data of the current obstacle object at each historical moment.

[0137] In an optional embodiment, the second anomaly detection module includes:

[0138] The first data acquisition unit is configured to acquire the first average driving value, the first standard deviation of the first driving data within the second preset time period before each historical moment, the second average driving value and the second standard deviation of the second driving data within the second preset time period corresponding to the first driving data of the current obstacle object within each historical moment.

[0139] The first driving difference determination unit is configured to determine the first driving difference between the first driving data of the current obstacle object at each historical time and the first driving average value.

[0140] The second driving difference determination unit is configured to determine a second driving difference between the second driving data of the current obstacle object at each historical time and the second driving mean.

[0141] The first anomaly detection result determination unit is configured to perform an operation to determine the first anomaly detection result based on the first driving difference and the first standard deviation.

[0142] The second anomaly detection result determination unit is configured to perform an operation to determine the second anomaly detection result based on the second driving difference and the second standard deviation.

[0143] In an optional embodiment, the second anomaly detection module includes:

[0144] The second data acquisition unit is configured to acquire the first driving data of the current obstacle object within the third preset time period corresponding to each historical moment and the second driving data within the third preset time period, wherein the third preset time period corresponding to each historical moment includes each historical moment.

[0145] The anomaly detection unit is configured to input the first driving data within the third preset time period and the second driving data within the third preset time period into a preset anomaly detection model to perform anomaly detection, and obtain the first anomaly detection result and the second anomaly detection result.

[0146] In an optional embodiment, the blind spot warning processing module 440 includes:

[0147] The current fusion driving data acquisition unit is configured to acquire the current fusion driving data of the current vehicle at the current moment when the probability of the current obstacle object entering the blind spot at the preset time is less than the first preset threshold, as indicated by the blind spot warning recognition result.

[0148] The vehicle accident recognition unit is configured to input the current fused driving data and the current environmental data into the vehicle accident recognition model to perform vehicle accident recognition and obtain vehicle accident recognition results.

[0149] The first preset warning information feedback unit is configured to feed back the first preset warning information when the vehicle accident identification result indicates that the probability of the current vehicle being involved in an accident is greater than or equal to a second preset threshold.

[0150] In an optional embodiment, the blind spot warning processing module 440 includes:

[0151] The second preset warning information feedback unit is configured to feed back second preset warning information when the blind zone warning identification result indicates that the probability of the current obstacle object entering the blind zone at the preset time is greater than or equal to the first preset threshold.

[0152] In an optional embodiment, the location prediction module 420 includes:

[0153] The motion mode determination unit is configured to determine the motion mode corresponding to the current obstacle object based on the first fused driving data;

[0154] The position prediction unit is configured to perform position prediction processing on the current obstacle object based on the motion mode and the first fused driving data to obtain the predicted position information.

[0155] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0156] Figure 5This is a block diagram of an electronic device for vehicle blind spot warning provided in an embodiment of this application. The electronic device can be a terminal, and its internal structure diagram can be as follows. Figure 5 As shown, the electronic device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a vehicle blind spot warning method. The display screen can be a liquid crystal display (LCD) or an e-ink display. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse.

[0157] Figure 6 This is a block diagram of another electronic device for vehicle blind spot warning provided in an embodiment of this application. The electronic device can be a server, and its internal structure diagram can be as follows: Figure 6 As shown, the electronic device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a vehicle blind spot warning method.

[0158] Those skilled in the art will understand that Figure 5 or Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present disclosure and does not constitute a limitation on the electronic device to which the present disclosure is applied. A specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0159] In an exemplary embodiment, an electronic device is also provided, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement a vehicle blind spot warning method as described in the embodiments of this disclosure.

[0160] In an exemplary embodiment, a computer-readable storage medium is also provided, wherein when the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the vehicle blind spot warning method of the present disclosure embodiments.

[0161] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the vehicle blind spot warning method provided in the various optional implementations described above.

[0162] It is understood that in the specific implementation of this application, user-related data is involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0163] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0164] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0165] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A vehicle blind spot warning method, characterized in that, The method includes: When a blind spot occurs during the current vehicle's driving, the system acquires first fused driving data, current environmental data, and the location information of the blind spot corresponding to the current obstacle object within a preset area corresponding to the current vehicle. The first fused driving data includes fused driving data of the current obstacle object at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing first filtered data and second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering the first driving data of the current obstacle object at each historical moment for abnormal driving data, and the second filtered data is obtained by filtering the second driving data of the current obstacle object at each historical moment for abnormal driving data. The first weight and the second weight are determined based on the number of occurrences of abnormal driving data of the current obstacle object within a first preset time period before each historical moment. The first driving data is driving data extracted from images collected by a camera device, and the second driving data is driving data collected by a preset sensor. Based on the first fused driving data, the current obstacle object is subjected to position prediction processing to obtain the predicted position information of the current obstacle object after a preset time after the current moment; The predicted location information, the regional location information, and the current environmental data are input into the blind spot early warning identification model to perform blind spot early warning identification and obtain the blind spot early warning identification result. Based on the blind spot warning identification results, blind spot warning processing is performed on the current vehicle.

2. The method according to claim 1, characterized in that, The method further includes: Anomaly detection is performed on the first fused driving data to determine the third anomaly detection result; The step of performing position prediction processing on the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object within a second preset time period after the current time includes: If the third anomaly detection result indicates that there is no abnormal driving data in the first fused driving data, the step of performing position prediction processing on the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object within a second preset time period after the current time is executed.

3. The method according to claim 2, characterized in that, The method further includes: If the third anomaly detection result indicates that there is abnormal driving data in the first fused driving data, the first fused driving data is subjected to anomaly filtering processing based on the third anomaly detection result to obtain the second fused driving data. Based on the interpolation method, the second fused driving data is corrected to obtain the third fused driving data; The step of performing position prediction processing on the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object within a second preset time period after the current time includes: Based on the third fused driving data, the current obstacle object is subjected to position prediction processing to obtain the predicted position information.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Obtain the first driving data of the current obstacle object at each historical moment and the second driving data of the current obstacle object at each historical moment; Anomaly detection is performed on the first driving data and the second driving data respectively to determine the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data; Based on the first anomaly detection result and the second anomaly detection result, the first driving data and the second driving data are respectively subjected to anomaly data filtering processing to obtain the first filtered data and the second filtered data. Based on the first and second anomaly detection results of the current obstacle object within the first preset time period before each historical moment, a first occurrence count and a second occurrence count are determined respectively. The first occurrence count is the number of times abnormal driving data of the current obstacle object appears in the first driving data within the first preset time period before each historical moment; the second occurrence count is the number of times abnormal driving data of the current obstacle object appears in the second driving data within the first preset time period before each historical moment. Based on the first occurrence count and the second occurrence count, the first weight and the second weight are determined. The first weight represents the effectiveness of the first filtered data, and the second weight represents the effectiveness of the second filtered data. The first weight is negatively correlated with the first occurrence coefficient and positively correlated with the second occurrence count; the second weight is positively correlated with the first occurrence coefficient and negatively correlated with the second occurrence count. Based on the first weight and the second weight, the first filtered data and the second filtered data are fused to obtain the fused driving data of the current obstacle object at each historical moment.

5. The method according to claim 4, characterized in that, The step of performing anomaly detection on the first driving data and the second driving data respectively, and determining the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data, includes: The first driving mean, the first standard deviation, the second driving mean, and the second standard deviation of the driving data within the second preset time period before each historical moment are obtained for the current obstacle object. Determine the first driving difference between the first driving data of the current obstacle object at each historical moment and the first driving average; Determine the second driving difference between the second driving data of the current obstacle object at each historical moment and the second driving mean; The first anomaly detection result is determined based on the first driving difference and the first standard deviation; The second anomaly detection result is determined based on the second driving difference and the second standard deviation.

6. The method according to claim 4, characterized in that, The step of performing anomaly detection on the first driving data and the second driving data respectively, and determining the first anomaly detection result corresponding to the first driving data and the second anomaly detection result corresponding to the second driving data, includes: Obtain the first driving data and the second driving data of the current obstacle object within the third preset time period corresponding to each historical moment, wherein the third preset time period corresponding to each historical moment includes each historical moment; The first driving data within the third preset time period and the second driving data within the third preset time period are input into a preset anomaly identification model for anomaly identification, and the first anomaly detection result and the second anomaly detection result are obtained.

7. The method according to any one of claims 1 to 3, characterized in that, The blind spot warning processing for the current vehicle based on the blind spot warning recognition result includes: If the blind spot warning recognition result indicates that the probability of the current obstacle entering the blind spot at the preset time is less than the first preset threshold, the current fused driving data of the current vehicle at the current moment is obtained; The current fused driving data and the current environmental data are input into the vehicle accident recognition model to perform vehicle accident recognition and obtain the vehicle accident recognition result. If the vehicle accident identification result indicates that the probability of the current vehicle being involved in an accident is greater than or equal to a second preset threshold, a first preset warning message is fed back.

8. The method according to any one of claims 1 to 3, characterized in that, The blind spot warning processing for the current vehicle based on the blind spot warning recognition result includes: If the blind spot warning identification result indicates that the probability of the current obstacle object entering the blind spot within the preset time is greater than or equal to the first preset threshold, a second preset warning message is fed back.

9. The method according to any one of claims 1 to 3, characterized in that, The step of performing position prediction processing on the current obstacle object based on the first fused driving data to obtain the predicted position information of the current obstacle object within a second preset time period after the current time includes: Based on the first fused driving data, the motion pattern corresponding to the current obstacle object is determined; Based on the motion pattern and the first fused driving data, the current obstacle object is subjected to position prediction processing to obtain the predicted position information.

10. A vehicle blind spot warning device, characterized in that, The device includes: The first data acquisition module is configured to, when a blind spot occurs during the current vehicle's driving, acquire first fused driving data, current environmental data, and the area location information corresponding to the blind spot within a preset area corresponding to the current vehicle. The first fused driving data includes fused driving data of the current obstacle at each historical moment within a historical time period. The fused driving data at each historical moment is obtained by fusing first filtered data and second filtered data based on a first weight and a second weight. The first filtered data is obtained by filtering the first driving data of the current obstacle at each historical moment for abnormal driving data, and the second filtered data is obtained by filtering the second driving data of the current obstacle at each historical moment for abnormal driving data. The first weight and the second weight are determined based on the number of occurrences of abnormal driving data of the current obstacle within a first preset time period before each historical moment. The first driving data is driving data extracted from images collected by a camera device, and the second driving data is driving data collected by a preset sensor. The location prediction module is configured to perform location prediction processing on the current obstacle object based on the first fused driving data, and obtain the predicted location information of the current obstacle object after a preset time after the current time. The blind spot warning and identification module is configured to input the predicted location information, the area location information and the current environmental data into the blind spot warning and identification model to perform blind spot warning and identification, and obtain the blind spot warning and identification result. The blind spot warning processing module is configured to perform blind spot warning processing on the current vehicle based on the blind spot warning recognition result.

11. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the vehicle blind spot warning method as described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to perform the vehicle blind spot warning method as described in any one of claims 1 to 9.

Images