A collision warning method, system, device and computer readable storage medium

Abstract

This application discloses a collision warning method, system, device, and computer-readable storage medium, relating to the field of image processing technology. The method involves acquiring an environmental image of the environment in which a target object, including vehicles or robots, is located; partitioning the environmental image to obtain target partition images; determining the target grayscale value of the target partition image; searching for the target individual corresponding to the target partition image in a pre-set individual database based on the target grayscale value, the individual database storing the known grayscale values ​​of a first number of individuals; acquiring the target motion trend of the target object and the target individual within a target time period; and determining the target collision probability value between the target object and the target individual in a pre-set collision database based on the target motion trend, the collision database storing the collision probability values ​​corresponding to a second number of known motion trends. This application achieves high flexibility by determining the target collision probability value through image processing and database matching methods.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and more particularly, to a collision warning method, system, device and computer readable storage medium. BACKGROUND

[0002] Currently, target objects such as vehicles and robots may collide with the outside world during movement. In order to avoid such situations, radars and sensors can be installed in target objects such as vehicles and robots to analyze the movement environment of the target objects to avoid collisions. However, the way in which radars and sensors guide target objects to avoid collisions is relatively mechanical, not flexible enough, and has poor applicability.

[0003] To sum up, how to improve the flexibility of the collision warning method is a problem that needs to be solved by the technical personnel in the field at present. SUMMARY

[0004] The purpose of the present application is to provide a collision warning method which can solve the technical problem of how to improve the flexibility of the collision warning method to some extent. The present application also provides a collision warning system, device and computer readable storage medium.

[0005] In a first aspect, the present application provides a collision warning method, comprising:

[0006] obtaining an environment image of an environment in which a target object is located, the target object comprising a vehicle or a robot;

[0007] partitioning the environment image to obtain a target partition image;

[0008] determining a target gray value of the target partition image;

[0009] in a preset individual database, based on the target gray value, searching for a target individual corresponding to the target partition image, the individual database storing known gray values of a first number of individuals;

[0010] obtaining a target motion trend of the target object and the target individual within a target time period;

[0011] in a preset collision database, based on the target motion trend, determining a target collision probability value between the target object and the target individual, the collision database storing collision probability values corresponding to a second number of known motion trends.

[0012] In an exemplary embodiment, the partitioning the environment image to obtain a target partition image comprises:

[0013] performing nine equal divisions on the environment image to obtain nine candidate partitions;

[0014] In the second, fourth, sixth and eighth of the candidate partitions, a corresponding rectangular region image is selected as the target partition image;

[0015] In the fifth of the candidate partitions, a corresponding elliptical region image is selected as the target partition image;

[0016] The counting rule of the candidate partitions is from left to right and from top to bottom.

[0017] In an example embodiment, the searching, in the preset individual database, for a target individual corresponding to the target partition image based on the target gray value includes:

[0018] The known gray values of the to-be-matched individuals in the individual database are sorted to obtain sorted gray values;

[0019] The sorted gray values are split into a third number of sorted sub-gray values;

[0020] The bisection method is used to search for, in all the sorted sub-gray values, a matching gray value having a difference from the target gray value within a target range;

[0021] When a first matching gray value is found, the to-be-matched individual corresponding to the first matching gray value is determined as the target individual, and the searching process is ended.

[0022] In an example embodiment, the determining, in the preset collision database, of a target collision probability value between the target object and the target individual based on the target motion trend includes:

[0023] The known motion trends in the collision database are obtained;

[0024] Particle swarm parameters are initialized based on the known motion trends;

[0025] The positions and velocities of each particle are randomly initialized based on the similarity between the known motion trends and the target motion trend;

[0026] It is determined whether the optimal particles in two consecutive times are the same and whether the iteration number exceeds a preset value;

[0027] If the optimal particles in two consecutive times are the same and / or the iteration number exceeds the preset value, the known motion trend corresponding to the optimal particle is determined as a matching motion trend, and the collision probability value corresponding to the matching motion trend is determined as the target collision probability value;

[0028] If the optimal particles are not the same in two consecutive times and the iteration number does not exceed the preset value, the speed and position of each particle are updated, the fitness value of each particle is calculated, the individual historical optimal fitness value and position of each particle are updated, the group historical optimal fitness value and position are updated, other parameters including the inertia weight and the iteration number are updated, and the step of judging whether the optimal particles in two consecutive times are the same and the subsequent steps are returned.

[0029] In an example embodiment, after determining the target collision probability value between the target object and the target individual based on the target motion trend, the method further comprises:

[0030] generating a corresponding collision prompt voice based on the target collision probability value;

[0031] playing the collision prompt voice based on a playing device of the target object.

[0032] In an example embodiment, the step of generating a corresponding collision prompt voice based on the target collision probability value comprises:

[0033] analyzing the type of the target individual;

[0034] if the type of the target individual is a static individual, generating the collision prompt voice prompting the target object to avoid collision;

[0035] if the type of the target individual is a dynamic individual, generating the collision prompt voice prompting the target object and the target individual to avoid collision.

[0036] In an example embodiment, the step of obtaining the target motion trend of the target object and the target individual in a target period comprises:

[0037] obtaining a relative position image of the target object and the target individual in the target period;

[0038] analyzing the relative position image in time sequence to obtain the target motion trend, wherein the target motion trend comprises relative speed, direction, relative distance, and trend change factor.

[0039] In a second aspect, the application provides a collision warning system, comprising:

[0040] a first acquisition module configured to acquire an environment image of an environment in which a target object is located, wherein the target object comprises a vehicle or a robot;

[0041] a first partition module configured to partition the environment image to obtain a target partition image;

[0042] The first determining module is configured to determine a target gray value of the target partition image;

[0043] The first searching module is configured to search, in a preset individual database, a target individual corresponding to the target partition image based on the target gray value, the individual database storing known gray values of a first number of individuals;

[0044] The second acquiring module is configured to acquire a target motion trend of the target object and the target individual within a target period;

[0045] The second determining module is configured to determine, in a preset collision database, a target collision probability value between the target object and the target individual based on the target motion trend, the collision database storing collision probability values corresponding to a second number of known motion trends.

[0046] In a third aspect, the present application provides a collision warning device, comprising:

[0047] The memory is configured to store a computer program;

[0048] The processor is configured to implement the steps of the collision warning method according to any one of the above aspects when executing the computer program.

[0049] In a fourth aspect, the present application provides a computer readable storage medium, the computer readable storage medium storing a computer program, the computer program being executed by a processor to implement the steps of the collision warning method according to any one of the above aspects.

[0050] The collision warning method provided in the application obtains an environment image of an environment where a target object is located, the target object including a vehicle or a robot; partitions the environment image to obtain a target partition image; determines a target gray value of the target partition image; in a preset individual database, finds a target individual corresponding to the target partition image based on the target gray value, the known gray values of a first number of individuals being stored in the individual database; obtains a target motion trend of the target object and the target individual within a target period; in a preset collision database, determines a target collision probability value between the target object and the target individual based on the target motion trend, the collision probability values corresponding to a second number of known motion trends being stored in the collision database. In the application, the environment image of the target object is partitioned to obtain the target partition image of the target individual that can reflect the possibility of collision with the target object, and the target individual can be determined in the individual database through gray value matching, so that the target individual is determined through the image processing method, the determination flexibility of the target individual is improved, and then the target collision probability value between the target object and the target individual is determined in the collision database through the target motion trend between the target object and the target individual and motion trend matching, so that the mechanical collision prediction of the radar and the sensor is avoided, and the flexibility is high. The collision warning system, the device and the computer readable storage medium provided in the application also solve the corresponding technical problems. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only constitute the embodiments of the application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of the provided drawings.

[0052] Figure 1 The first flow chart of the collision warning method provided in the embodiments of the application;

[0053] Figure 2 The second flow chart of the collision warning method provided in the embodiments of the application;

[0054] Figure 3 The schematic diagram of the environment image;

[0055] Figure 4 The schematic diagram of the target partition image;

[0056] Figure 5 The third flow chart of the collision warning method provided in the embodiments of the application;

[0057] Figure 6 The fourth flow chart of the collision warning method provided in the embodiments of the application;

[0058] Figure 7 A structure diagram of a collision warning system provided by an embodiment of the present application is shown in FIG. 1.

[0059] Figure 8 A structure diagram of a collision warning device provided by an embodiment of the present application is shown in FIG. 2.

[0060] Figure 9 Another structure diagram of a collision warning device provided by an embodiment of the present application is shown in FIG. 3. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0062] Please refer to Figure 1 , Figure 1 A first flowchart of a collision warning method provided by an embodiment of the present application is shown in FIG. 4.

[0063] The collision warning method provided by the embodiments of the present application can include the following steps.

[0064] Step S101: Obtain an environment image of an environment where a target object is located, the target object including a vehicle or a robot.

[0065] In actual application, the environment image of the environment where the target object is located can be obtained first, for example, the environment image of the environment where the target object is located can be obtained by means of an image acquisition device of the target object. Since the present application is to analyze collision, the environment image can be an environment image around the target object, for example, one or more of an environment image in front of the target object, an environment image behind the target object, an environment image on the left of the target object, and an environment image on the right of the target object, etc., which is not specifically limited herein.

[0066] It can be understood that, since the collision occurs to a moving object, the target object in the present application refers to a moving object, for example, a car, a robot, an airplane, etc., the type of which can be determined according to a specific application scenario.

[0067] Step S102: Partition the environment image to obtain a target partition image.

[0068] Step S103: Determine a target gray value of the target partition image.

[0069] Step S104: In the preset individual database, the target individual corresponding to the target partition image is searched based on the target gray value. The known gray values of the first number of individuals are stored in the individual database.

[0070] In actual application, because the collision is caused by two objects, after the environment image of the environment where the target object is located is acquired, the target individual that can collide with the target object in the environment image is analyzed. In the present application, the target individual is determined by a gray value matching method, that is, the environment image is partitioned to obtain a target partition image. Because each target partition image carries part of the information of the target individual, the feature quantity of the target individual can be increased by partitioning. Thus, a plurality of target gray values corresponding to a plurality of target partition images can be determined subsequently, and then the target individual can be accurately searched in the preset individual database based on the target gray value.

[0071] It can be understood that the known gray values of the first number of individuals are stored in the individual database in the present application. Thus, when a certain known gray value matches the target gray value, the individual corresponding to the known gray value can be determined as the target individual. It should be noted that the type of the target individual can be determined according to actual needs, such as a wall, a car, a pedestrian, an electric vehicle, etc.

[0072] Step S105: The target motion trend of the target object and the target individual in the target time period is acquired.

[0073] In actual application, after the target individual corresponding to the target partition image is searched in the preset individual database based on the target gray value, the collision probability value between the target object and the target individual can be predicted. Because the collision is caused by motion, the target object and the target individual can only collide when the motion trend between them meets the condition. Therefore, the target motion trend of the target object and the target individual in the target time period can be acquired, and the collision probability value between the target object and the target individual is analyzed based on the target motion trend.

[0074] In a specific application scenario, in the process of acquiring the target motion trend of the target object and the target individual in the target time period, in order to flexibly and conveniently determine the target motion trend of the target object and the target individual in the target time period, the relative position image of the target object and the target individual in the target time period can be acquired; the relative position image is analyzed in time sequence to obtain the target motion trend, that is, the target motion trend can be determined by image analysis method, and the type of the target motion trend can be determined according to actual needs, such as relative speed, direction, relative distance, trend change factor, etc. The trend change factor can be the data type with the largest change amplitude in the relative speed, direction, and relative distance, etc. The present application does not make specific limitation here.

[0075] Step S106: determining a target collision probability value between the target object and the target individual based on the target motion trend in a preset collision database, the collision database storing collision probability values corresponding to a second number of known motion trends.

[0076] In actual application, after determining the target motion trend, the target collision probability value can be quickly determined in the preset collision database through motion trend matching, so as to convert the dynamic collision prediction into a static data searching and matching process, and enhance the accuracy of the collision prediction.

[0077] It should be noted that the collision database involved in the present application stores a second number of known motion trends and corresponding collision probability values, the types of which can be determined according to actual needs, such as the collision probability values of two cars under different motion trends; in addition, the motion trends of two objects that have collided can be analyzed and sorted to generate the collision database, and of course there can be other ways to generate the collision database, which are not limited in the present application.

[0078] The collision warning method provided by the present application obtains an environment image of an environment in which a target object is located, the target object including a vehicle or a robot; partitions the environment image to obtain a target partition image; determines a target gray value of the target partition image; searches for a target individual corresponding to the target partition image based on the target gray value in a preset individual database, the individual database storing known gray values of a first number of individuals; obtains a target motion trend of the target object and the target individual within a target period; and determines a target collision probability value between the target object and the target individual based on the target motion trend in a preset collision database, the collision database storing collision probability values corresponding to a second number of known motion trends. In the present application, the environment image of the target object is partitioned to obtain a target partition image that can reflect a target individual that has a possibility of collision with the target object, and the target individual can be determined in the individual database through gray value matching, so that the target individual is determined through an image processing method, and the determination flexibility of the target individual is improved. Thereafter, the target motion trend between the target object and the target individual is collected, and the target collision probability value between the target object and the target individual is determined in the collision database through motion trend matching, so that the mechanical nature of the radar and the sensor for collision prediction is eliminated, and the flexibility is high.

[0079] Please refer to Figure 2 , Figure 2 A second flowchart of a collision warning method provided by an embodiment of the present application.

[0080] The collision warning method provided by the embodiment of the present application can include the following steps:

[0081] Step S201: Obtain an environment image of an environment where a target object is located, the target object including a vehicle or a robot.

[0082] Step S202: Divide the environment image into nine equal parts to obtain nine candidate sub-zones.

[0083] Step S203: In the second, fourth, sixth, and eighth candidate sub-zones, a corresponding rectangular region image is selected as a target sub-zone image.

[0084] Step S204: In the fifth candidate sub-zone, a corresponding elliptical region image is selected as a target sub-zone image; wherein the counting rule of the candidate sub-zones is from left to right and from top to bottom.

[0085] In actual applications, in the process of dividing the environment image to obtain the target sub-zone image, in order to facilitate operation, the environment image can be divided into nine equal parts to obtain nine candidate sub-zones; in the second, fourth, sixth, and eighth candidate sub-zones, a corresponding rectangular region image is selected as a target sub-zone image; in the fifth candidate sub-zone, a corresponding elliptical region image is selected as a target sub-zone image, as shown in Figure 3 and Figure 4 , wherein the target individual is a truck, and the white area is the target sub-zone image; wherein the counting rule of the candidate sub-zones is from left to right and from top to bottom.

[0086] It can be understood that in the process of dividing the environment image to obtain the target sub-zone image, one-third of the width of the environment image and one-third of the height of the environment image can be determined, and the center of the environment image is taken as the origin, one-third of the width is taken as the major axis of the ellipse, and one-third of the height is taken as the minor axis of the ellipse to divide the elliptical region image as the target sub-zone image, and on both sides of the elliptical region image, the minor axis is taken as the height of the rectangle to divide the rectangular region image as the target sub-zone image, and on both sides of the elliptical region image, the major axis is taken as the width of the rectangle to divide the rectangular region image as the target sub-zone image, and so on to quickly divide the environment image into five target sub-zone images.

[0087] It can be understood that in specific application scenarios, the way and the number of dividing the environment image into target sub-zone images can be flexibly determined as needed, which is not limited in the present application.

[0088] Step S205: Determine the target gray value of the target sub-zone image.

[0089] Step S206: In the preset individual database, a target individual corresponding to the target partition image is searched based on the target gray value, and the known gray values of the first number of individuals are stored in the individual database.

[0090] Step S207: A target motion trend of the target object and the target individual in a target period is acquired.

[0091] Step S208: In the preset collision database, a target collision probability value between the target object and the target individual is determined based on the target motion trend, and the collision probability values corresponding to the second number of known motion trends are stored in the collision database.

[0092] Please refer to Figure 5 , Figure 5 A third flowchart of a collision warning method provided by the embodiment of the application.

[0093] The collision warning method provided by the embodiment of the application can include the following steps:

[0094] Step S301: An environment image of an environment where a target object is located is acquired, and the target object includes a vehicle or a robot.

[0095] Step S302: The environment image is partitioned to obtain a target partition image.

[0096] Step S303: A target gray value of the target partition image is determined.

[0097] Step S304: In the preset individual database, the known gray values of the to-be-matched individuals in the individual database are sorted to obtain sorted gray values.

[0098] Step S305: The sorted gray values are split into a third number of sorted sub-gray values.

[0099] Step S306: The bisection search method is used to search for matching gray values in all the sorted sub-gray values, and the difference between the target gray value and the matching gray value is within a target range.

[0100] Step S307: When the first matching gray value is found, the to-be-matched individual corresponding to the first matching gray value is determined as the target individual, and the searching process is ended.

[0101] In actual application, in the process of searching the target individual corresponding to the target partition image based on the target gray value in the preset individual database, the known gray values of the to-be-matched individuals in the individual database can be sorted first to obtain sorted gray values; the sorted gray values can be then split into a third number of sorted sub-gray values, such as four sorted sub-gray values, etc.; finally, a matching gray value whose difference with the target gray value is within the target range is searched in all the sorted sub-gray values by the dichotomy searching method, and when the first matching gray value is found, the to-be-matched individual corresponding to the first matching gray value is determined as the target individual, and the searching process is ended, so as to quickly search the target individual in the individual database based on the dichotomy searching method. Of course, a matching gray value whose difference with the target gray value is within the target range can also be searched in each sorted sub-gray value, and finally the to-be-matched individual corresponding to the matching gray value with the smallest difference or the most occurrences can be selected as the target individual, which is not limited in the present application.

[0102] It can be understood that the category of the target individual searched subsequently can be different according to different ways of determining the target gray value of the target partition image, that is, the way of determining the target gray value of the target partition image can be determined according to different application requirements, such as the sum of the gray values of each target partition image can be taken as the target gray value, the average of the gray values of all target partition images can be taken as the target gray value, etc.

[0103] Step S308: acquiring a target motion trend of the target object and the target individual in a target period.

[0104] Step S309: determining a target collision probability value between the target object and the target individual based on the target motion trend in a preset collision database, and the collision database stores collision probability values corresponding to a second number of known motion trends.

[0105] Please refer to Figure 6 , Figure 6 A fourth flowchart of a collision warning method provided by the embodiments of the present application.

[0106] The collision warning method provided by the embodiments of the present application can include the following steps:

[0107] Step S401: acquiring an environment image of an environment where a target object is located, and the target object includes a vehicle or a robot.

[0108] Step S402: partitioning the environment image to obtain a target partition image.

[0109] Step S403: determining a target gray value of the target partition image.

[0110] Step S404: In the preset individual database, a target individual corresponding to the target partition image is searched based on the target gray value, and the known gray values of the first number of individuals are stored in the individual database.

[0111] Step S405: A target motion trend of the target object and the target individual in the target period is acquired.

[0112] Step S406: A known motion trend in the collision database is acquired.

[0113] Step S407: Particle swarm parameters are initialized based on the known motion trend.

[0114] Step S408: The position and the speed of each particle are randomly initialized based on the similarity between the known motion trend and the target motion trend.

[0115] Step S409: Whether the optimal particles in two consecutive times are the same and whether the iteration number exceeds a preset value are judged; if the optimal particles in two consecutive times are the same and / or the iteration number exceeds the preset value, step S410 is executed; if the optimal particles in two consecutive times are not the same and the iteration number does not exceed the preset value, step S411 is executed.

[0116] Step S410: The known motion trend corresponding to the optimal particle is determined as a matching motion trend, and the collision probability value corresponding to the matching motion trend is determined as a target collision probability value.

[0117] Step S411: The speed and the position of each particle are updated; the fitness value of each particle is calculated; the individual historical optimal fitness value and the position of each particle are updated; the group historical optimal fitness value and the position are updated; other parameters including the inertia weight and the iteration number are updated; and the step of judging whether the optimal particles in two consecutive times are the same and the subsequent steps are returned.

[0118] In actual application, in the process of determining the target collision probability value between the target object and the target individual based on the target motion trend in the preset collision database, the particle swarm algorithm can be used to quickly find the target collision probability value from the collision database, that is, the known motion trend in the collision database can be obtained; the particle swarm parameters are initialized based on the known motion trend; the position and speed of each particle are randomly initialized based on the similarity between the known motion trend and the target motion trend; it is judged whether the optimal particles of two consecutive times are the same, and whether the iteration number exceeds the preset value, wherein the preset value can be determined according to experience, etc.; if the optimal particles of two consecutive times are the same and / or the iteration number exceeds the preset value, the known motion trend corresponding to the optimal particle is determined as the matching motion trend, and the collision probability value corresponding to the matching motion trend is determined as the target collision probability value; if the optimal particles of two consecutive times are not the same and the iteration number does not exceed the preset value, the speed and position of each particle are updated; the fitness value of each particle is calculated; the individual historical optimal fitness value and position of each particle are updated; the group historical optimal fitness value and position are updated; other parameters including inertia weight, iteration number are updated; the steps of judging whether the optimal particles of two consecutive times are the same and the subsequent steps are returned.

[0119] It should be noted that the convergence condition in the present embodiment is that the optimal particles of two consecutive times are the same and / or the iteration number exceeds the preset value. Since the optimal particles of two consecutive times are the same, the particle swarm algorithm falls into local convergence, and when the iteration number exceeds the preset value, the particle swarm algorithm falls into over-convergence. Therefore, the present application can avoid local convergence and over-convergence in the search of the collision probability value, and ensure the search effect of the collision probability value.

[0120] In actual application, after determining the target collision probability value between the target object and the target individual based on the target motion trend, in order to avoid collision between the target object and the target individual, a corresponding collision prompt voice can also be generated based on the target collision probability value; the collision prompt voice is played based on the playing device of the target object.

[0121] In a specific application scenario, in the process of generating a corresponding collision prompt voice based on the target collision probability value, the type of the target individual can be analyzed; if the type of the target individual is a static individual, a collision prompt voice prompting the target object to avoid collision is generated, such as in the case where the target object is a car and the target individual is a wall, a collision prompt voice prompting the driver of the car to avoid collision can be generated; if the type of the target individual is a dynamic individual, a collision prompt voice prompting the target object and the target individual to avoid collision is generated, such as in the case where the target object is a car and the target object is an electric vehicle, a collision prompt voice prompting the drivers of the car and the electric vehicle to avoid collision is generated, etc.

[0122] Please refer to Figure 7 ,Figure 7 A structure diagram of a collision warning system provided by an embodiment of the present application.

[0123] The collision warning system provided by an embodiment of the present application can include:

[0124] The first acquisition module 101 is configured to acquire an environment image of an environment in which a target object is located, the target object including a vehicle or a robot.

[0125] The first partition module 102 is configured to partition the environment image to obtain a target partition image.

[0126] The first determination module 103 is configured to determine a target gray value of the target partition image.

[0127] The first search module 104 is configured to search for a target individual corresponding to the target partition image based on the target gray value in a preset individual database, the individual database storing known gray values of a first number of individuals.

[0128] The second acquisition module 105 is configured to acquire a target motion trend of the target object and the target individual within a target period.

[0129] The second determination module 106 is configured to determine a target collision probability value between the target object and the target individual based on the target motion trend in a preset collision database, the collision database storing collision probability values corresponding to a second number of known motion trends.

[0130] The collision warning system provided by an embodiment of the present application can include:

[0131] The first partition unit is configured to divide the environment image into nine equal parts to obtain nine to-be-selected partitions.

[0132] The first selection unit is configured to select a corresponding rectangular region image as the target partition image in the second to-be-selected partition, the fourth to-be-selected partition, the sixth to-be-selected partition, and the eighth to-be-selected partition.

[0133] The second selection unit is configured to select a corresponding elliptical region image as the target partition image in the fifth to-be-selected partition.

[0134] The counting rule of the to-be-selected partitions is from left to right and from top to bottom.

[0135] The collision warning system provided by an embodiment of the present application can include:

[0136] The first sorting unit is configured to sort the known gray values of the to-be-matched individuals in the individual database to obtain sorted gray values.

[0137] The first splitting unit is configured to split the ordered gray value into a third number of ordered sub-gray values.

[0138] The first searching unit is configured to search for a matching gray value in all the ordered sub-gray values by using a binary search method, and the matching gray value has a difference from the target gray value within the target range.

[0139] The first processing unit is configured to determine the to-be-matched individual corresponding to the first matching gray value as the target individual when the first matching gray value is found, and end the searching process.

[0140] The second determining module can be specifically configured to: acquire a known motion trend in a collision database; initialize particle swarm parameters based on the known motion trend; randomly initialize a position and a speed of each particle based on a similarity between the known motion trend and a target motion trend; judge whether the optimal particles in two consecutive times are the same and whether an iteration number exceeds a preset value; if the optimal particles in two consecutive times are the same and / or the iteration number exceeds the preset value, determine the known motion trend corresponding to the optimal particle as a matching motion trend, and determine a collision probability value corresponding to the matching motion trend as a target collision probability value; if the optimal particles in two consecutive times are not the same and the iteration number does not exceed the preset value, update the speed and the position of each particle; calculate an adaptive value of each particle; update an individual historical optimal adaptive value and a position of each particle; update a group historical optimal adaptive value and a position; update other parameters, the other parameters including an inertia weight and the iteration number; and return to execute the step of judging whether the optimal particles in two consecutive times are the same and the subsequent steps.

[0141] The collision warning system provided in the embodiment of the present application can further include:

[0142] The first generating module is configured to generate a corresponding collision prompt voice based on the target collision probability value after the second determining module determines the target collision probability value between the target object and the target individual based on the target motion trend; and play the collision prompt voice based on a playing device of the target object.

[0143] The collision warning system provided in the embodiment of the present application can further include:

[0144] The first generating unit is configured to analyze a type of the target individual; if the type of the target individual is a static individual, generate a collision prompt voice for prompting the target object to avoid collision; and if the type of the target individual is a dynamic individual, generate a collision prompt voice for prompting the target object and the target individual to avoid collision.

[0145] The collision warning system provided in the embodiment of the present application can further include:

[0146] The first acquisition unit is configured to acquire a relative position image of the target object and the target individual in a target period.

[0147] The first analysis unit is configured to analyze the relative position image in time sequence to obtain a target motion trend, wherein the target motion trend comprises a relative speed, a direction, a relative distance, and a trend change factor.

[0148] The present application also provides a collision warning device and a computer readable storage medium, both of which have the corresponding effects of the collision warning method provided by the embodiments of the present application. Please refer to Figure 8 , Figure 8 The collision warning device provided by the embodiments of the present application is shown in the structural diagram.

[0149] The collision warning device provided by the embodiments of the present application comprises a memory 201 and a processor 202. The memory 201 stores a computer program. When the processor 202 executes the computer program, the steps of the collision warning method described in any of the above embodiments are implemented.

[0150] Please refer to Figure 9 , the other collision warning device provided by the embodiments of the present application can further comprise: an input port 203 connected with the processor 202, configured to transmit an external input command to the processor 202; a display unit 204 connected with the processor 202, configured to display the processing result of the processor 202 to the outside world; and a communication module 205 connected with the processor 202, configured to realize the communication between the collision warning device and the outside world. The display unit 204 can be a display panel, a laser scanning display, etc. The communication mode adopted by the communication module 205 includes but is not limited to mobile high-definition link technology (HML), universal serial bus (USB), high-definition multimedia interface (HDMI), wireless connection: wireless fidelity technology (WiFi), Bluetooth communication technology, low-power Bluetooth communication technology, and communication technology based on IEEE 802.11s.

[0151] The computer readable storage medium provided by the embodiments of the present application stores a computer program. When the computer program is executed by a processor, the steps of the collision warning method described in any of the above embodiments are implemented.

[0152] The computer readable storage medium involved in the present application includes random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the technical field.

[0153] The collision warning system, the device and the computer readable storage medium provided in the embodiments of the present application are described in detail in the corresponding part of the collision warning method provided in the embodiments of the present application. Details are not described here. In addition, the part of the above technical solutions provided in the embodiments of the present application which is consistent with the implementation principle of the corresponding technical solution in the prior art is not described in detail, so as not to be too verbose.

[0154] It should also be noted that, in this document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0155] The above description of disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A collision warning method characterized by, The method comprises the following steps: acquiring an environment image of an environment in which a target object is located, the target object comprising a vehicle or a robot; partitioning the environment image to obtain a target partition image; determining a target gray value of the target partition image; in a preset individual database, searching for a target individual corresponding to the target partition image based on the target gray value, the individual database storing known gray values of a first number of individuals; acquiring a target motion trend of the target object and the target individual within a target period; in a preset collision database, determining a target collision probability value between the target object and the target individual based on the target motion trend, the collision database storing collision probability values corresponding to a second number of known motion trends; wherein the target motion trend comprises relative speed, direction, relative distance, and trend change factor; and the determination of the target collision probability value between the target object and the target individual based on the target motion trend in the preset collision database comprises: acquiring the known motion trends in the collision database; initializing particle swarm parameters based on the known motion trends; randomly initializing the position and speed of each particle based on the similarity between the known motion trends and the target motion trend; judging whether the optimal particles in two consecutive times are the same and whether the iteration number exceeds a preset value; if the optimal particles in two consecutive times are the same and / or the iteration number exceeds the preset value, determining the known motion trend corresponding to the optimal particle as a matching motion trend and determining the collision probability value corresponding to the matching motion trend as the target collision probability value; if the optimal particles in two consecutive times are not the same and the iteration number does not exceed the preset value, updating the speed and position of each particle, calculating the fitness value of each particle, updating the individual historical optimal fitness value and position of each particle, updating the group historical optimal fitness value and position, updating other parameters including the inertia weight and the iteration number, and returning to the step of judging whether the optimal particles in two consecutive times are the same and the subsequent steps.

2. The method of claim 1, wherein, The partitioning of the environment image to obtain a target partition image comprises: ninth equal division of the environment image to obtain nine candidate partitions; selecting a corresponding rectangular region image as the target partition image in the second, fourth, sixth, and eighth candidate partitions; selecting a corresponding elliptical region image as the target partition image in the fifth candidate partition; wherein the counting rule of the candidate partitions is from left to right and from top to bottom.

3. The method of claim 1, wherein, The searching for a target individual corresponding to the target partition image based on the target gray value in the preset individual database comprises: sorting the known gray values of the individuals to be matched in the individual database to obtain sorted gray values; splitting the sorted gray values into a third number of sorted sub-gray values; finding a matching gray value in all the sorting sub-gray values by dichotomy method, wherein the difference between the target gray value and the matching gray value is within a target range; when the first matching gray value is found, determining the first matching gray value corresponding to the to-be-matched individual as the target individual, and ending the finding process.

4. The method of claim 1, wherein, After determining the target collision probability value between the target object and the target individual based on the target motion trend, the method further comprises: generating a corresponding collision prompt voice based on the target collision probability value; playing the collision prompt voice based on a playing device of the target object.

5. The method of claim 4, wherein, The method of generating a corresponding collision prompt voice based on the target collision probability value comprises: analyzing the type of the target individual; if the type of the target individual is a static individual, generating the collision prompt voice for prompting the target object to avoid collision; if the type of the target individual is a dynamic individual, generating the collision prompt voice for prompting the target object and the target individual to avoid collision.

6. The method of claim 1, wherein, The method of obtaining the target motion trend of the target object and the target individual within a target time period comprises: obtaining a relative position image of the target object and the target individual within the target time period; analyzing the relative position image in time sequence to obtain the target motion trend, wherein the target motion trend comprises relative speed, direction, relative distance, and trend change factor.

7. A collision warning system characterized by comprising: The method comprises: a first obtaining module configured to obtain an environment image of an environment in which a target object is located, wherein the target object comprises a vehicle or a robot; a first partitioning module configured to partition the environment image to obtain a target partition image; a first determining module configured to determine a target gray value of the target partition image; a first finding module configured to find a target individual corresponding to the target partition image based on the target gray value in a preset individual database, wherein known gray values of a first number of individuals are stored in the individual database; a second obtaining module configured to obtain a target motion trend of the target object and the target individual within a target time period; a second determining module configured to determine a target collision probability value between the target object and the target individual based on the target motion trend in a preset collision database, wherein collision probability values corresponding to a second number of known motion trends are stored in the collision database; wherein the target motion trend comprises relative speed, direction, relative distance, and trend change factor. The second determining module is configured to: acquire the known motion trend in the collision database; initialize particle swarm parameters based on the known motion trend; randomly initialize the position and speed of each particle based on the similarity between the known motion trend and the target motion trend; determine whether the optimal particles in two consecutive times are the same and whether the iteration number exceeds a preset value; if the optimal particles in two consecutive times are the same and / or the iteration number exceeds the preset value, determine the known motion trend corresponding to the optimal particle as a matching motion trend, and determine the collision probability value corresponding to the matching motion trend as the target collision probability value; if the optimal particles in two consecutive times are not the same and the iteration number does not exceed the preset value, update the speed and position of each particle; calculate the fitness value of each particle; update the individual historical optimal fitness value and position of each particle; update the group historical optimal fitness value and position; update other parameters, the other parameters including an inertia weight, the iteration number; and return to perform the step of determining whether the optimal particles in two consecutive times are the same and the subsequent steps.

8. A collision warning apparatus characterized by comprising: Comprise: a memory for storing a computer program; a processor for implementing the steps of the collision warning method according to any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium, characterized in that, The computer program stored in the computer readable storage medium is executed by the processor to implement the steps of the collision warning method according to any one of claims 1 to 6.

Images