Safety warning method, device and computer-readable storage medium for blind spot of vision

Abstract

The present application discloses a safety warning method, device and computer-readable storage medium for a blind spot of vision. The method includes: after obtaining a current frame image, obtaining a first set corresponding to a historical frame image adjacent to the current frame image, wherein both the current frame image and the historical frame image include a target blind spot of vision that cannot be observed by a driver, and the first set includes first targets located in a first area in the historical frame image, wherein the first area covers the target blind spot of vision and extends outside the target blind spot of vision; determining first positions of the respective first targets in the first set in the current frame image; determining whether the respective first targets enter the target blind spot of vision in the current frame image according to the first positions corresponding to the respective first targets; and in the case of determining that at least one first target enters the target blind spot of vision in the current frame image, performing a first alarm prompt inside the target vehicle. The method of the present application can reduce the occurrence of accidents.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular, to a safety warning method, device, and computer-readable storage medium for a blind spot of vision. Background Art

[0002] During driving, some vehicles or pedestrians may cut in front of a large truck. Due to the high front of the large truck, there will be a blind spot of vision near the front of the truck. When a vehicle or pedestrian falls into the blind spot of vision near the front of the truck, the driver cannot perceive the vehicle or pedestrian that has fallen into the blind spot of vision, which may cause vehicle collisions or even pedestrians to be involved under the vehicle, resulting in life-threatening situations. Summary of the Invention

[0003] This application provides a safety warning method, device, and computer-readable storage medium for a blind spot of vision, which can improve driving safety and reduce the occurrence of safety accidents.

[0004] In a first aspect of an embodiment of this application, a safety warning method for a blind spot of vision is provided. The method includes: after obtaining a current frame image, obtaining a first set corresponding to a historical frame image adjacent to the current frame image, where both the current frame image and the historical frame image include a target blind spot of vision that cannot be observed by a driver, and the first set includes first targets in a first area in the historical frame image, where the first area covers the target blind spot of vision and extends outside the target blind spot of vision; determining first positions of the first targets in the first set in the current frame image; determining whether each of the first targets enters the target blind spot of vision in the current frame image according to the first positions corresponding to the first targets; and in the case of determining that at least one of the first targets enters the target blind spot of vision in the current frame image, performing a first alarm prompt inside the target vehicle.

[0005] In a second aspect of an embodiment of this application, a safety warning device is provided. The safety warning device includes a processor, a memory, and a communication circuit. The processor is respectively coupled to the memory and the communication circuit. Program data is stored in the memory, and the processor realizes the steps in any one of the above methods by executing the program data in the memory.

[0006] In a third aspect of an embodiment of this application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and the computer program can be executed by a processor to realize the steps in any one of the above methods.

[0007] The beneficial effects are as follows: Considering the targets that can enter the blind area of the target field of view in the current frame image, it is highly probable or basically in the first area in the historical frame adjacent to the current frame, and it is less probable or basically not in the area outside the first area. Therefore, compared with the prior art solution of first determining the positions of all targets in the current frame image and then judging whether there are targets entering the blind area of the target field of view based on the positions of each target, in this embodiment, it is only necessary to judge whether each target in the first set enters the blind area of the target field of view, so as to narrow the search range, improve the efficiency of judging whether there are targets entering the blind area of the target field of view, and further reduce the occurrence of accidents and improve driving safety. Description of the Drawings

[0008] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings, where:

[0009] Figure 1 is a schematic flowchart of an embodiment of the safety warning method for the blind area of the field of view of the present application;

[0010] Figure 2 is a schematic diagram of the division structure of each area in the current frame image and the historical frame image adjacent to the current frame image;

[0011] Figure 3 is Figure 1 a schematic flowchart of step S130 in

[0012] Figure 4 is Figure 1 a schematic diagram of another part of the flowchart of the embodiment;

[0013] Figure 5 is Figure 4 a schematic flowchart of step S180 in

[0014] Figure 6 is a schematic diagram of the structure of an embodiment of the safety warning system of the present application;

[0015] Figure 7 is a schematic diagram of the structure of an embodiment of the safety warning device of the present application;

[0016] Figure 8 is a schematic diagram of the structure of another embodiment of the safety warning device of the present application;

[0017] Figure 9 is a schematic diagram of the structure of an embodiment of the computer-readable storage medium of the present application. Detailed Embodiments

[0018] Next, in combination with the accompanying drawings in the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0019] First of all, it should be noted that the safety warning method for the blind area of vision in the present application is not only applicable to large vehicles such as trucks and lorries, but also can be used for small vehicles such as cars. In short, it can be applicable to any vehicle with a blind area of vision.

[0020] At the same time, the safety warning method for the blind area in the present application is executed by a safety warning device, which is installed on the target vehicle and gives a safety warning to the driver during the driving of the target vehicle.

[0021] Refer to Figure 1 , Figure 1 which is a schematic flowchart of an implementation manner of the safety warning method for the blind area of vision in the application. The method includes:

[0022] S110: After obtaining the current frame image, obtain a first set corresponding to the historical frame image adjacent to the current frame image. Among them, both the current frame image and the historical frame image include the target blind area of vision that cannot be observed by the driver of the target vehicle. The first set includes a first target in the first area in the historical frame image, where the first area covers the target blind area of vision and extends to the outside of the target blind area of vision.

[0023] Specifically, a camera device is installed outside the target vehicle. During the driving of the target vehicle, the camera device takes pictures of the blind area of vision that cannot be observed by the driver of the target vehicle. At the same time, the safety warning device establishes a communication connection with the camera device and receives the images taken by the camera device.

[0024] Among them, the blind area of vision is the area that cannot be observed by the driver of the target vehicle during the driving of the target vehicle. When people, vehicles or animals enter this blind area of vision, there is a risk of collision with the target vehicle.

[0025] Among them, after the safety warning device receives the image sent by the camera device, it analyzes and processes the image. Among them, the current frame image is the image that the safety warning device is about to analyze and process currently, and the historical frame image adjacent to the current frame image is the image that the safety warning device has analyzed and processed most recently. It can be understood that the shooting time point corresponding to the historical frame image adjacent to the current frame image is earlier than the shooting time point corresponding to the current frame image.

[0026] For example, after the camera device captures the blind spot of the target vehicle's field of vision, it forms a video stream and sends it to the safety warning device. After receiving the video stream, if the safety warning device analyzes and processes each frame of the video stream, the frame number of the current frame image in the video stream is i, and the frame number of the historical frame image adjacent to the current frame image in the video stream is i-1. If the safety warning device does not analyze and process each frame of the video stream, but analyzes and processes the images in the video stream according to a preset frame interval, then the frame number of the current frame image in the video stream is i, and the frame number of the historical frame image adjacent to the current frame image in the video stream is iN, where N is the preset frame interval.

[0027] Among them, the current frame image and the historical frame image both include the target blind spot of the target vehicle, that is, the current frame image and the historical frame image both include the same blind spot of the target vehicle. The technical problem to be solved by this application is how to provide a safety warning to the driver for the target blind spot of the target vehicle during the driving process of the target vehicle.

[0028] The camera device may simultaneously photograph multiple blind areas of the target vehicle, or may only photograph one blind area of ​​the target vehicle. It is understandable that when the camera device photographs multiple blind areas, multiple blind areas exist in the image received by the safety warning device, and when the safety warning device photographs only one blind area, only one blind area exists in the image received by the safety warning device.

[0029] Among them, when the camera device simultaneously shoots multiple blind spots, the current frame image and the historical frame image can simultaneously include multiple target blind spots on the target vehicle. At this time, steps S110 to S140 are respectively executed for the multiple target blind spots on the target vehicle. For example, the current frame image and the historical frame image both include the first target blind spot, the second target blind spot and the third target blind spot on the target vehicle (for example, the first target blind spot is the blind spot on the left side of the front of the vehicle, the second target blind spot is the blind spot on the right side of the front of the vehicle, and the third target blind spot is the blind spot on the rear side of the vehicle). At this time, steps S110 to S140 are respectively executed for the first target blind spot, the second target blind spot and the third target blind spot, so that during the driving process of the target vehicle, a safety warning can be given to the driver for multiple blind spots of the target vehicle at the same time.

[0030] For the sake of convenience, the following description is based on the assumption that the camera device only shoots one target blind spot of the target vehicle, that is, the current frame image and the historical frame image only include one target blind spot.

[0031] In an application scenario, after installing the imaging device on the target vehicle, the installation position and shooting angle of the imaging device remain unchanged all the time. Then, first, the target field of view blind area is calibrated in the image captured by the imaging device, and in any subsequent image captured by the imaging device, this calibrated area is used as the target field of view blind area.

[0032] At the same time, combined with Figure 2 , both the current frame image and the historical frame image include the target field of view blind area and the first area. Among them, the target field of view blind area is represented by 101, and the first area is represented by 102. At the same time, the first area covers the target field of view blind area and extends to the outside of the target field of view blind area.

[0033] At the same time, the first set includes all first targets in the first area in the historical frame image. Among them, all first targets in the first area can include any objects such as all people, vehicles, and animals in the first area that may collide with the target vehicle and cause accidents.

[0034] Among them, the first set can specifically include the unique identifiers of all first targets in the first area in the historical frame image, such as ID numbers.

[0035] S120: Determine the first positions of each first target in the first set in the current frame image.

[0036] Specifically, after determining the first set in step S110, determine the positions of each first target in the first set in the current frame image respectively, that is, obtain the corresponding first positions of each first target.

[0037] Among them, step S120 can specifically be: first determine the positions of all targets in the current frame image, and then, among the positions of all targets, find the positions of each first target in the first set, that is, obtain the corresponding first positions of each first target.

[0038] S130: According to the first positions corresponding to each first target, determine whether each first target enters the target field of view blind area in the current frame image.

[0039] Specifically, according to the first positions corresponding to each first target, it can be determined which first targets enter the target field of view blind area.

[0040] S140: In the case of determining that at least one first target enters the target field of view blind area in the current frame image, perform a first alarm prompt inside the target vehicle.

[0041] Specifically, if one first target enters the target field of view blind area in the current frame image, perform a first alarm prompt for the driver inside the target vehicle, and ask the driver to pay attention to avoiding the first target that enters the target field of view blind area.

[0042] Among them, the first alarm prompt includes but is not limited to at least one of a sound prompt, a light prompt, and a picture prompt, and specifically can be any prompt method that can play a warning role.

[0043] In this embodiment, considering the targets that can enter the target field of view blind area in the current frame image, it is highly probable or basically in the first area in the historical frame adjacent to the current frame, and it is less probable or basically not in the area outside the first area. Therefore, compared with the prior art solution of first determining the positions of all targets in the current frame image and then respectively judging whether there are targets entering the target field of view blind area according to the positions of each target, in this embodiment, it is only necessary to judge whether each target in the first set enters the target field of view blind area, so as to narrow the search range, improve the efficiency of judging whether there are targets entering the target field of view blind area, and further reduce the occurrence of accidents and improve driving safety.

[0044] Among them, after step S140, the method in this embodiment further includes: constructing and saving the first set corresponding to the current frame image, and the first set includes all targets in the first area in the current frame image. Specifically, saving the first set corresponding to the current frame image can improve the efficiency of the subsequent safety warning device when analyzing and processing the next frame image adjacent to the current frame image.

[0045] Continue to refer to Figure 2 , in this embodiment, the target field of view blind area and the first area are both rectangular. And in an application scenario, assuming that the upper left vertex coordinates of the target field of view blind area are (x11, y11), the lower right vertex coordinates are (x12, y12), the upper left vertex coordinates of the first area are (x21, y21), and the lower right vertex coordinates are (x22, y22), then there is the following relationship:

[0046] x21 = x11 - a; y21 = y11 + a; x22 = x12 + a; y22 = y12 - a.

[0047] It should be noted that this application does not limit the shape, size, etc. of the target field of view blind area and the first area, as long as the first area covers the target field of view blind area and extends to the outside of the first area. For example, the target field of view blind area is rectangular, and the first area is a circle that covers the target field of view blind area and extends to the outside of the first area.

[0048] Refer to Figure 3 , in this embodiment, the step of step S130 for judging whether each first target enters the target field of view blind area in the current frame image includes:

[0049] S131: Determine the coordinates of the first vertex and the second vertex of the target field of view blind area, where the first vertex and the second vertex are on the diagonal of the target field of view blind area.

[0050] In an application scenario, the first vertex and the second vertex are respectively the upper left vertex and the lower right vertex of the target field of view blind area.

[0051] In another application scenario, the first vertex and the second vertex are respectively the lower right vertex and the upper left vertex of the target field of view blind area.

[0052] S132: Determine whether each first target enters the target field of view blind area in the current frame image respectively according to the coordinates of each first target center point, the coordinates of the first vertex, and the coordinates of the second vertex.

[0053] In an application scenario, for any first target, determine whether the abscissa of its center is between the abscissa of the first vertex and the abscissa of the second vertex, and whether the ordinate of the center is between the ordinate of the first vertex and the ordinate of the second vertex. If both conditions are satisfied, it is determined that the first target enters the target field of view blind area in the current frame image, otherwise it does not enter the target field of view blind area.

[0054] In another application scenario, for any first target, if the coordinates (xi, yi) of its center point satisfy the following conditions, it is determined that it enters the target field of view blind area in the current frame image:

[0055] First, record the coordinates of the first vertex as (x11, y11), and the coordinates of the second vertex as (x12, y12);

[0056] Then determine the values of V1 and V2 according to the following formula: V1 = (xi - x11) × (x12 - xi); V2 = (yi - y12) × (y11 - yi);

[0057] Finally, if both V1 and V2 are greater than or equal to 0, it is determined that the first target enters the target field of view blind area, otherwise it is determined that the first target does not enter the target field of view blind area.

[0058] That is to say, first determine the first difference between the abscissa of the first target center point and the abscissa of the first vertex, the second difference between the abscissa of the second vertex and the abscissa of the first target center point, the third difference between the ordinate of the first target center point and the ordinate of the second vertex, and the fourth difference between the ordinate of the second vertex and the ordinate of the first target center point; then determine the first product of the corresponding first difference and the second difference of the first target, and the second product of the corresponding third difference and the fourth difference; finally, if both the first product and the second product corresponding to the first target are greater than or equal to zero, it is determined that the first target enters the target field of view blind area, otherwise it is determined that the first target does not enter the target field of view blind area.

[0059] In other embodiments, it is also possible to separately determine the intersection over union (IoU) between the area occupied by the detection box of each first target and the target vision blind area. If the IoU is greater than the IoU threshold, it is determined that the first target has entered the target vision blind area; otherwise, it is determined that the first target has not entered the target vision blind area.

[0060] In this embodiment, when it is determined in step S140 that at least one first target has entered the target vision blind area in the current frame image, step S140 further includes:

[0061] (a) Separately count the total duration for each first target that has entered the target vision blind area to enter the target vision blind area within the most recent preset time period.

[0062] Specifically, if it is determined through the previous steps that a first target has entered the target vision blind area in the current frame image, then separately count the total duration for each first target that has entered the target vision blind area within the most recent preset time period, such as the most recent 1 minute or 5 minutes.

[0063] It can be understood that since counting the total duration for each first target to enter the target vision blind area within the most recent preset time period requires using historical information, each time the safety warning device determines that a first target in the first set has entered the target vision blind area, it is necessary to correspondingly save the shooting time point corresponding to the current frame image with the first target. Thus, based on the saved shooting time point, the total duration for which the first target has entered the target vision blind area within the preset time period can be determined.

[0064] (b) In response to the total duration corresponding to the first target exceeding the duration threshold, give an alarm prompt for the first target outside the target vehicle.

[0065] Specifically, if the total duration corresponding to the first target exceeds the duration threshold, it means that after multiple reminders to the driver, the first target is still in the target vision blind area. At this time, simply reminding the driver is far from enough, and it is also necessary to give a reminder for the first target outside the target vehicle so that it actively stays away from the target vision blind area of the target vehicle.

[0066] Among them, the method of giving an alarm prompt for the first target outside the target vehicle can be any method such as lighting or sound, which is not limited here.

[0067] Among them, when giving an alarm prompt for the first target, the information in the prompt can also carry the characteristic information of the first target. For example, if the first target is a vehicle, the information in the prompt can carry the license plate information of the vehicle; if the first target is a person, the information in the prompt can carry information such as the color of the person's clothes and the position relative to the target vehicle, so that the first target can timely walk out of the target vision blind area of the target vehicle.

[0068] In this embodiment, when it is determined in step S140 that at least one first target enters the target vision blind area in the current frame image, it further includes:

[0069] (c) Determine the relative positions of the respective first targets that enter the target vision blind area and the target vehicle.

[0070] Specifically, according to the position of the target vision blind area relative to the target vehicle and the positions of the respective first targets that enter the target vision blind area within the target vision blind area, the relative positions of the respective first targets that enter the target vision blind area and the target vehicle can be determined.

[0071] (d) Plan the movement trajectory of the target vehicle according to the relative positions corresponding to the respective first targets.

[0072] Specifically, according to the positions of the respective first targets relative to the target vehicle, the movement trajectory of the target vehicle can be planned, so that the target vehicle travels according to the planned movement trajectory, and all the first targets that enter the target vision blind area can be made to no longer be in the target vision blind area as quickly as possible.

[0073] (e) Send the planned movement trajectory to the driver.

[0074] Specifically, after planning the movement trajectory, send the planned movement trajectory to the driver for the driver's reference.

[0075] Among them, the planned movement trajectory can be sent to the driver's mobile phone (the mobile phone is pre - established with a communication connection with the safety warning device), or it can be sent to the console of the target vehicle and then displayed through the display screen of the console. In short, no matter by what means it is sent to the driver, as long as the driver can see it.

[0076] Refer to Figure 4 , the method of this embodiment further includes:

[0077] S150: Obtain a second set corresponding to the historical frame image, where the second set includes second targets in a second area in the historical frame image, and the second area covers the first area and extends outside the first area.

[0078] Among them, obtaining the second set corresponding to the historical frame can be carried out synchronously with obtaining the first set corresponding to the historical frame, or can be carried out after obtaining the first set corresponding to the historical frame.

[0079] S160: Determine the second positions of the respective second targets in the second set in the current frame image.

[0080] S170: According to the second positions corresponding to the respective second targets, determine whether the respective second targets enter a first target area that extends from the first area to the outside of the target field of view blind area in the current frame image.

[0081] Among them, step S170 may specifically include: for each second target, first determine whether it enters the first area in the current frame image. If the determination result is that it enters the first area, then further determine whether it enters the target field of view blind area. If it does not enter the target field of view blind area, it is determined that it enters the first target area.

[0082] Among them, the method for determining whether the second target enters the first area is the same as the method for determining whether the first target enters the target field of view blind area described above. For specific details, please refer to the relevant content above and will not be elaborated here.

[0083] S180: When it is determined that at least one second target enters the first target area in the current frame image, perform a second alarm prompt inside the target vehicle.

[0084] Among them, the warning effect of the first alarm prompt is higher than that of the second alarm prompt.

[0085] In an application scenario, step S180 is executed after step S140.

[0086] Specifically, in combination with Figure 2 , the relative positional relationship between the second area and the first area is similar to the relative positional relationship between the first area and the target field of view blind area. Among them, Figure 2 the second area is denoted by reference numeral 103 in

[0087] Meanwhile, the specific processes of steps S150 to S180 are similar to the specific processes of steps S110 to S180 described above. For specific details, please refer to the relevant content above and will not be elaborated here.

[0088] Among them, it can be understood that the first target area that extends from the first area to the outside of the target field of view blind area is specifically the area where the first area does not overlap with the target field of view blind area.

[0089] In this embodiment, considering the targets that can enter the first target area in the current frame image, it is highly probable or basically in the second area in the historical frames adjacent to the current frame, and it is less probable or basically not in the area outside the second area. Therefore, in this embodiment, as long as it is determined whether each target in the second set enters the first target area, it can be quickly determined whether there is a target entering the first target area in the current frame image, and when a target enters, a prompt is given to the driver.

[0090] At the same time, because the severity of the existence of targets in the target field of view blind area is greater than the severity of the existence of targets in the first target area, the warning effect of the second alarm prompt can be less than the warning effect of the first alarm prompt. For example, both the first alarm prompt and the second alarm prompt are sound alarm prompts, but the volume of the first alarm prompt is greater than the volume of the second alarm prompt.

[0091] Among them, considering that if the first alarm prompt and the second alarm prompt are given simultaneously, it may make the driver at a loss. Therefore, in this embodiment, there may also be a prerequisite for executing step S150, specifically: step S140 determines that none of the first targets enter the target field of view blind area in the current frame image. That is, only when all the first targets do not enter the target field of view blind area in the current frame image, step S150 is executed, otherwise step S150 is not executed.

[0092] Or in other embodiments, step S180 specifically includes: in response to at least one second target entering the first target area in the current frame image, determining whether the first alarm prompt is being given currently; in response to the first alarm prompt not being given currently, giving a second alarm prompt inside the target vehicle.

[0093] Specifically, if it is determined that there is a second target entering the first target area in the current frame image, it is determined whether the first alarm prompt is being given currently. If the first alarm prompt is being given currently, the second alarm prompt is not given. Only when the first alarm prompt is not being given currently, the second alarm prompt is given.

[0094] In this embodiment, when a target enters the first target area, a prompt is given to the driver, which can play a warning role, give the driver sufficient preparation time, and avoid in advance the targets that may enter the target field of view blind area, further reducing the occurrence of accidents.

[0095] Among them, similar to the first set corresponding to the above-mentioned preservation of the current frame image, for the analysis and processing of the next frame image, after step S180, it further includes: constructing and saving a second set corresponding to the current frame image, where the second set includes all targets in the second region in the current frame image. Specifically, saving the second set corresponding to the current frame image can improve the efficiency of the subsequent safety warning device when analyzing and processing the next frame image adjacent to the current frame image.

[0096] In this embodiment, referring to Figure 5 , step S180 specifically includes: S181: When it is determined that at least one second target enters the first target area in the current frame image, estimate the movement directions of the respective second targets that enter the first target area to obtain the estimated movement directions corresponding to the respective second targets.

[0097] S182: When it is determined that the estimated movement direction corresponding to at least one second target is towards the target vision blind area, give a second alarm prompt inside the target vehicle.

[0098] Specifically, if the second target enters the first target area in the current frame image but moves away from the target vehicle, the probability that it enters the target vision blind area in the next frame image is relatively low. Therefore, in order to improve the accuracy of the alarm, step S181 also estimates the movement directions of the respective second targets that enter the first target area to obtain the estimated movement directions of the respective second targets. As long as the estimated movement direction of one second target is towards the target vision blind area, a second alarm prompt will be given; otherwise, no second alarm prompt will be given. That is to say, if the estimated movement directions corresponding to all second targets that enter the first target area are away from the target vehicle, no second alarm prompt will be given.

[0099] Among them, the step of estimating the movement direction of the second target can be: determining the estimated movement direction of the second target vehicle according to the positions of the second target in multiple historical frame images. For example, if it is determined according to the positions of the second target in multiple historical frame images that the position of the second target is getting closer and closer to the target vision blind area, then determine that the estimated movement direction of the second target is towards the target vision blind area. Or, if the second target is a vehicle, the estimated movement direction of the vehicle can also be determined according to whether the front of the vehicle is towards the target vision blind area or away from the target vision blind area.

[0100] Among them, referring to the setting methods of the second region and the first region, and continuing to combine Figure 2 , a third region ( Figure 2 denoted by reference numeral 104 in

[0101] (A) Obtain a third set corresponding to a historical frame image, where the third set includes third targets in a third area in the historical frame image, and the third area covers the second area and extends outside the second area.

[0102] (B) Determine the third positions of the respective third targets in the third set in the current frame image.

[0103] (C) Based on the third positions corresponding to the respective third targets, determine whether the respective third targets enter a second target area that extends from the second area to the outside of the first area in the current frame image.

[0104] (D) In response to at least one third target entering the second target area in the current frame image, perform a third alarm prompt inside the target vehicle.

[0105] Wherein, the warning effect of the second alarm prompt is higher than that of the third alarm prompt.

[0106] Wherein, the specific processes of steps A to D are similar to the specific processes of the above steps S150 to S180. For details, refer to the above relevant content and will not be elaborated here.

[0107] In an application scenario, the prerequisite for executing step (A) is that all second targets do not enter the first target area in the current frame image.

[0108] And so on, more areas such as a fourth area and a fifth area can also be set and will not be elaborated in detail here.

[0109] Refer to Figure 6 , Figure 6 is a schematic structural diagram of an implementation manner of the safety warning system of the present application. The safety warning system 200 includes a photographing device 210 and a safety warning device 220. The photographing device 210 is communicatively connected to the safety warning device 220.

[0110] Wherein, the photographing device 210 is installed outside the target vehicle and photographs the target vision blind area of the target vehicle.

[0111] The safety warning device 220 receives the image photographed by the photographing device 210 and analyzes and processes the image. The analysis and processing process includes:

[0112] (a) After obtaining the current frame image, obtain a first set corresponding to a historical frame image adjacent to the current frame image.

[0113] Among them, both the current frame image and the historical frame image include a target blind spot that cannot be observed by the driver of the target vehicle, a first area that covers the target blind spot and extends outside the target blind spot, and a second area that covers the first area and extends outside the first area.

[0114] Among them, the first set includes first targets in the first area in the historical frame image.

[0115] (b) Determine the first positions of the respective first targets in the first set in the current frame image.

[0116] (c) Determine whether each first target enters the target blind spot in the current frame image according to the corresponding first position of each first target.

[0117] (d) In the case where it is determined that at least one first target enters the target blind spot in the current frame image, perform a first alarm prompt inside the target vehicle.

[0118] In an application scenario, after step (d), it further includes:

[0119] (e) In response to all first targets not entering the target blind spot in the current frame image, obtain a second set corresponding to the historical frame image adjacent to the current frame image.

[0120] Among them, the second set includes second targets in the second area in the historical frame image.

[0121] (f) Determine the second positions of the respective second targets in the second set in the current frame image.

[0122] (g) Determine whether each second target enters the first target area where the first area extends to the target blind spot in the current frame image according to the corresponding second position of each second target.

[0123] (h) In the case where it is determined that at least one second target enters the first target area in the current frame image, perform a second alarm prompt inside the target vehicle.

[0124] In an application scenario, after step (h), it further includes:

[0125] (i) In response to all second targets not entering the first target area in the current frame image, obtain a third set corresponding to the historical frame image adjacent to the current frame image.

[0126] Among them, the third set includes third targets in the third area in the historical frame image.

[0127] (j) Determine the third positions of the respective third targets in the third set in the current frame image.

[0128] (k) Determine whether each third target enters the second target area where the second area extends to the first area in the current frame image according to the third position corresponding to each third target.

[0129] (l) When it is determined that at least one second target enters the second target area in the current frame image, a third alarm prompt is made inside the target vehicle.

[0130] Among them, the warning effects of the first alarm prompt, the second alarm prompt, and the third alarm prompt gradually decrease.

[0131] Among them, for the specific working process of the safety warning device 220, reference can be made to the above relevant content, which will not be elaborated here.

[0132] Refer to Figure 7 , Figure 7 FIG. is a schematic structural diagram of an embodiment of the safety warning device of the present application. The safety warning device 300 includes a processor 310, a memory 320, and a communication circuit 330. The processor 310 is respectively coupled to the memory 320 and the communication circuit 330. Program data is stored in the memory 320. The processor 310 implements the method steps in any of the above embodiments by executing the program data in the memory 320. For the detailed steps, reference can be made to the above embodiments, which will not be elaborated here.

[0133] Refer to Figure 8 , Figure 8 FIG. is a schematic structural diagram of an embodiment of the safety warning device of the present application. The safety warning device 400 includes an acquisition module 410, a first determination module 420, a second determination module 430, and an alarm prompt module 440.

[0134] The acquisition module 410 is configured to acquire a first set corresponding to a historical frame image adjacent to the current frame image after acquiring the current frame image. Among them, both the current frame image and the historical frame image include a target blind area that cannot be observed by the driver of the target vehicle. The first set includes first targets in the first area in the historical frame image. Among them, the first area covers the target blind area and extends to the outside of the target blind area.

[0135] The first determination module 420 is connected to the acquisition module 410 and is configured to determine the first position of each first target in the first set in the current frame image.

[0136] The second determination module 430 is connected to the first determination module 420 and is configured to determine whether each first target enters the target blind area in the current frame image according to the first position corresponding to each first target.

[0137] The alarm prompt module 440 is connected to the second determination module 430 and is configured to perform a first alarm prompt inside the target vehicle when it is determined that at least one first target enters the target vision blind area in the current frame image.

[0138] Wherein, the safety warning device 400 performs the method steps in any of the above embodiments during operation. For detailed steps, reference can be made to the relevant content above and will not be elaborated here.

[0139] Refer to Figure 9 , Figure 9 FIG. is a structural schematic diagram of an embodiment of the computer-readable storage medium of the present application. The computer-readable storage medium 500 stores a computer program 510, and the computer program 510 can be executed by a processor to implement the steps in any of the above methods.

[0140] Wherein, the computer-readable storage medium 500 can specifically be a device such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store the computer program 510, or it can also be a server storing the computer program 510. The server can send the stored computer program 510 to other devices for operation, or it can also run the stored computer program 510 by itself.

[0141] The above are only embodiments of the present application and do not limit the patent scope of the present application. Any equivalent structural or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A safety warning method for a blind spot of vision, characterized in that, the method includes: After obtaining the current frame image, obtain a first set corresponding to the historical frame image adjacent to the current frame image, wherein both the current frame image and the historical frame image include a target blind spot of vision that cannot be observed by the driver, and the first set includes a first target in a first area in the historical frame image, wherein the first area covers the target blind spot of vision and extends to the outside of the target blind spot of vision; Determine the first positions of the respective first targets in the first set in the current frame image; According to the respective first positions corresponding to the respective first targets, determine whether the respective first targets enter the target blind spot of vision in the current frame image; In the case where it is determined that at least one of the first targets enters the target blind spot of vision in the current frame image, perform a first alarm prompt inside the target vehicle; wherein, the method further includes: Obtain a second set corresponding to the historical frame image, the second set includes a second target in a second area in the historical frame image, wherein the second area covers the first area and extends to the outside of the first area; Determine the second positions of the respective second targets in the second set in the current frame image; According to the respective second positions corresponding to the respective second targets, determine whether the respective second targets enter a first target area that extends from the first area to the outside of the target blind spot of vision in the current frame image; In the case where it is determined that at least one of the second targets enters the first target area in the current frame image, perform a second alarm prompt inside the target vehicle; wherein the warning effect of the first alarm prompt is higher than that of the second alarm prompt.

2. The method according to claim 1, characterized in that, the step of performing a second alarm prompt inside the target vehicle in the case where it is determined that at least one of the second targets enters the first target area in the current frame image includes: In the case where it is determined that at least one of the second targets enters the first target area in the current frame image, respectively estimate the movement directions of the respective second targets that enter the first target area to obtain the estimated movement directions corresponding to the respective second targets; In the case where it is determined that the estimated movement direction corresponding to at least one of the second targets is towards the target blind spot of vision, perform the second alarm prompt inside the target vehicle.

3. The method according to claim 1, characterized in that, after performing the first alarm prompt and performing the second alarm prompt, further includes: Construct and save the first set and the second set corresponding to the current frame image, wherein the first set includes all targets in the first area in the current frame image, and the second set includes all targets in the second area in the current frame image.

4. The method according to claim 1, characterized in that, The target field of view blind area is rectangular; the step of determining whether each of the first targets enters the target field of view blind area in the current frame image according to the first positions corresponding to the respective first targets includes: Determining the coordinates of a first vertex and the coordinates of a second vertex of the target field of view blind area, wherein the first vertex and the second vertex are on the diagonal of the target field of view blind area; Determining whether each of the first targets enters the target field of view blind area in the current frame image respectively according to the coordinates of the center point of each of the first targets, the coordinates of the first vertex, and the coordinates of the second vertex.

5. The method according to claim 4, wherein, The step of determining whether each of the first targets enters the target field of view blind area in the current frame image respectively according to the coordinates of the center point of each of the first targets, the coordinates of the first vertex, and the coordinates of the second vertex includes: Respectively determining a first difference between the abscissa of the center point of each of the first targets and the abscissa of the first vertex, a second difference between the abscissa of the second vertex and the abscissa of the center point of each of the first targets, a third difference between the ordinate of the center point of each of the first targets and the ordinate of the second vertex, and a fourth difference between the ordinate of the second vertex and the ordinate of the center point of each of the first targets; Respectively determining a first product of the corresponding first difference and the second difference of each of the first targets, and a second product of the corresponding third difference and the fourth difference; In response to the first product and the second product corresponding to the first target both being greater than or equal to zero, determining that the first target enters the target field of view blind area.

6. The method according to claim 1, wherein, When it is determined that at least one of the first targets enters the target field of view blind area in the current frame image, the method further includes: Respectively counting the total duration for which each of the first targets that enters the target field of view blind area enters the target field of view blind area within a recent preset time period; In response to the total duration corresponding to the first target exceeding a duration threshold, giving an alarm prompt for the first target outside the target vehicle.

7. The method according to claim 1, wherein, When it is determined that at least one of the first targets enters the target field of view blind area in the current frame image, the method further includes: When it is determined that at least one of the first targets enters the target field of view blind area in the current frame image, respectively determining the relative positions of each of the first targets that enters the target field of view blind area and the target vehicle; Planning the movement trajectory of the target vehicle according to the relative positions corresponding to the respective first targets; Sending the planned movement trajectory to the driver.

8. A safety warning device, wherein, The safety warning device includes a processor, a memory, and a communication circuit. The processor is respectively coupled to the memory and the communication circuit. Program data is stored in the memory, and the processor executes the program data in the memory to implement the steps in the method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, and the computer program can be executed by a processor to implement the steps in the method according to any one of claims 1-7.

Images