CMS collision prevention alarm system based on machine vision

Abstract

The utility model discloses a CMS collision prevention alarm system based on machine vision, and relates to the technical field of automobiles, the CMS collision prevention alarm system based on machine vision comprises a CMS system, the CMS system comprises an early warning unit comprising two display screens; the monitoring unit comprises two paths of cameras; the string adding device is electrically connected with the output end of the monitoring unit; the deserializer is electrically connected with the output end of the serializer; the SOC chip is electrically connected with the output end of the deserializer, the SOC chip comprises an AI video processing module, and the AI video processing module processes monitoring unit video data received by the SOC chip through an AI algorithm; the collision judgment module is used for processing the monitoring unit video data received by the SOC chip through an AI algorithm; the output end of the SOC chip is electrically connected with the early warning unit. According to the utility model, the motion trail of an object behind the vehicle is pre-judged in advance through the CMS collision prevention alarm system, and the risk of vehicle collision is avoided through effective early warning.

Description

Technical Field

[0001] The utility model relates to the technical field of automobiles, in particular to a CMS collision prevention warning system based on machine vision. Background Art

[0002] The Camera Monitor System (CMS) replaces the traditional optical exterior mirror with a combination of a camera and a monitor. The display mode is that the external camera captures the image and displays it on the display screen in the cockpit after processing.

[0003] When a car is driving on the road, the driver has a limited view of the rear of the vehicle, so there are blind spots that the driver cannot see. Although the electronic rearview mirror has a larger field of view than the physical rearview mirror, and the general CMS system has a blind spot detection system. However, the blind spot detection can only warn objects in the blind spot, but not objects outside the blind spot. If the object behind the car does not enter the blind spot, but the object is moving at high speed, it is still impossible to determine whether the high-speed moving object will collide with the moving car. To this end, we provide a CMS collision prevention warning system based on machine vision. Utility Model Content

[0004] The utility model provides a CMS collision prevention warning system based on machine vision, which solves the problem of warning high-speed moving objects.

[0005] To achieve the above-mentioned purpose, the utility model provides the following technical solutions: a CMS collision prevention warning system based on machine vision, including a CMS system, wherein the CMS system includes: an early warning unit including two display screens;

[0006] Monitoring unit, including two cameras;

[0007] The serial adder is electrically connected to the output terminals of the monitoring units respectively;

[0008] a deserializer, electrically connected to an output terminal of the serializer; and

[0009] A SOC chip is electrically connected to the output end of the deserializer, and the SOC chip includes: an AI video processing module, the AI ​​video processing module processes the monitoring unit video data received by the SOC chip; and

[0010] A collision judgment module, which processes the monitoring unit video data received by the SOC chip;

[0011] The output end of the SOC chip is electrically connected to the early warning unit, the AI ​​video processing module includes a detection unit, a marking unit and an image processing unit, and the collision judgment module receives information from the image processing unit.

[0012] The utility model has the following beneficial effects:

[0013] (1) By predicting the movement trajectory of objects behind the vehicle in advance and providing effective warning, the risk of vehicle collision can be avoided;

[0014] (2) Based on machine vision algorithm, the algorithm has high accuracy and efficiency;

[0015] (3) No increase in costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is the block diagram of the CMS system of the utility model;

[0017] Figure 2 This is a flowchart based on the AI ​​algorithm of the utility model;

[0018] Figure 3 This is a block diagram of the AI ​​video processing module of the utility model.

[0019] In the figure: 1. CMS system; 2. Early warning unit; 21. Display screen; 3. Monitoring unit; 31. Camera; 4. Serializer; 5. Deserializer; 6. SOC chip; 61. AI video processing module; 62. Collision judgment module. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the utility model to clearly and completely describe the technical solutions in the embodiments of the utility model. Obviously, the described embodiments are only part of the embodiments of the utility model, not all of the embodiments. Based on the embodiments in the utility model, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the utility model.

[0021] like Figure 1-3 As shown, the utility model provides a technical solution: a CMS collision prevention warning system based on machine vision, including a CMS system 1, the CMS system 1 includes: an early warning unit 2, including two display screens 21, and an alarm prompt is performed on the display screen 21;

[0022] The monitoring unit 3 includes two cameras 31, which are arranged on the vehicle to collect video data;

[0023] The serial adder 4 is electrically connected to the output terminals of the monitoring units 3 respectively;

[0024] The deserializer 5 is electrically connected to the output end of the serializer 4; and

[0025] The SOC chip 6 is electrically connected to the output end of the deserializer 5. The SOC chip 6 includes: an AI video processing module 61. The AI ​​video processing module 61 processes the video data of the monitoring unit 3 received by the SOC chip 6 through an AI algorithm, including converting the received video frame data image into a BGR format, extracting a grayscale image, Gaussian blurring the grayscale image, binarizing the image, and dilating the image twice. The AI ​​video processing module 61 includes: a detection unit, which performs target object recognition on the video data collected by the camera 31, and recognizes whether the target object is a motor vehicle, a non-motor vehicle, a pedestrian or other objects;

[0026] Marking units, numbering the target objects, and binding them to the properties of the target objects;

[0027] and an image processing unit to obtain the contour of the target object;

[0028] And, a collision judgment module 62, the collision judgment module 62 processes the video data of the monitoring unit 3 received by the SOC chip 6 through an AI algorithm, the collision judgment module 62 receives information from the image processing unit, calculates the speed and movement direction of the target object by comparing with the previous frame, tracks the motion of the target object based on the motion algorithm of the target object contour, and judges whether it will collide with the vehicle driven by the driver according to the calculation result of the motion tracking algorithm, and if a collision occurs, an alarm prompt is given on the display screen 21;

[0029] The output end of the SOC chip 6 is electrically connected to the early warning unit 2;

[0030] When judging whether there is a collision, first receive the video frame data, convert the image into BGR format, extract the grayscale image, and Gaussian blur the grayscale image. Then use image binarization and image secondary expansion to highlight the outline, detect the object and obtain the outline, then obtain the target object properties, compare with the previous frame to calculate the target object speed and movement direction, judge whether there will be a collision, and issue an alarm message if there is a collision.

[0031] It should be noted that, in this article, terms such as "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or apparatus.

[0032] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A CMS collision prevention warning system based on machine vision, comprising a CMS system (1), wherein the CMS system (1) comprises: An early warning unit (2), comprising two display screens (21); A monitoring unit (3), comprising two cameras (31); The string adder (4) is electrically connected to the output end of the monitoring unit (3) respectively; A deserializer (5) is electrically connected to an output end of the serializer (4); and The SOC chip (6) is electrically connected to the output end of the deserializer (5), and is characterized in that: the SOC chip (6) comprises: an AI video processing module (61), the AI ​​video processing module (61) processes the video data of the monitoring unit (3) received by the SOC chip (6); and A collision judgment module (62), wherein the collision judgment module (62) processes the video data of the monitoring unit (3) received by the SOC chip (6); The output end of the SOC chip (6) is electrically connected to the early warning unit (2); the AI ​​video processing module (61) comprises a detection unit, a marking unit and an image processing unit; and the collision judgment module (62) receives information from the image processing unit.

Images