[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.
[0018] see figure 1 , A reversing safety system based on ADAS analysis of the pull-back camera image, including a pull-back camera, a CPU data processing module, an early warning diagnosis module, an APP processing module, a human-computer interaction display module, and a human-computer interaction voice module.
[0019] The pull-back camera is used to obtain reversing image data and input the data to the CPU data processing module;
[0020] The CPU data processing module is used to send image data to the early warning diagnosis module in the form of each frame;
[0021] The early warning diagnosis module is used for recognizing image data into corresponding object model information and returning the model information to the CPU data processing module;
[0022] The CPU data processing module is used to send corresponding object model information to the APP processing module;
[0023] The human-computer interaction display module is used to display the object model information processed by the APP processing module, and the human-computer interaction voice module is used to convert the object model information processed by the APP processing module into a dangerous level voice signal for playback.
[0024] Further, see figure 2 , The early warning diagnosis module is used to determine the region of interest (ROI area) of the image by filtering and analyzing each frame of image data, and then use the adaboost iterative algorithm to extract the object model information. Specifically, first extract the picture features of the region of interest, and then compare the picture features with pre-existing object model information to identify and screen out the corresponding object model information and the coordinate position of the object model on the image.
[0025] In order to get a good recognition effect, it is necessary to traverse and filter multiple target small areas, such as 20X20, 36X36, 48X48, in short, do several traversals, and the recognition effect is the best.
[0026] In the present invention, the reversing picture data is subjected to complete ADAS analysis to obtain the corresponding object model information and the information is sent to the APP processing module through the CPU data processing module, and the APP processing module outputs the information to the man-machine after the corresponding logical operation The interactive display module displays, and the danger level prompt sound is played through the human-computer interactive voice module. The driver can perform corresponding operations according to the danger level prompt sound and the content displayed by the display module. TTS voice can be added if the corresponding object model type is determined. Broadcast, such as beware of pedestrians, please pay attention to reversing! Remind the driver to pay attention to rear safety in time, and at the same time draw the corresponding object module on the right side of the LCD display whether it is in the left rear, directly behind, and back of the car, or in all three situations, so that the driver can pay attention to check situation.
[0027] Since the current smart rearview mirrors that support the reversing image function are basically not installed at the loading factory, if you want to install such smart rearview mirrors on different types of car bodies, such as cars, SUVs, small trucks, etc. For large trucks, etc., at this time, in order to make the captured images have a better viewing angle range, the position of the rear pull camera needs to correspond to the corresponding vehicle body, for example: SUV is higher, the height of the camera from the ground is higher; If the car is shorter, the height of the camera from the ground is lower. In addition, the height of the camera from the ground will also affect the difference in the imaging range and distance of the rear camera. Even in the same car, the pull-back camera is installed in different positions of the vehicle, such as the middle position, or left or right, the imaging range and sense of distance will be different. Therefore, in the present invention, the above factors are comprehensively considered to determine a reasonable early warning and alarm range in the imaging range. Specifically, see image 3 The pull-back camera sets the vertical middle horizontal line of the formed image plane as an alarm line, which divides the image plane into an early warning monitoring area and an early warning alarm area. The early warning alarm area is the area closest to the rear of the vehicle. The line can move up and down the image plane. The user can move up and down to adjust the position of the warning line on the image plane according to the vehicle body, and can also expand or reduce the warning video range according to their driving level.
[0028] The following further explains the role of the alarm line in the present invention. Pull the camera back to display the acquired image plane on the man-machine interactive display module, and the alarm line will be displayed in the image plane. The alarm line divides the image plane into early warning monitoring areas and early warning alarms. When an object appears on the image plane and falls into the early warning monitoring area, the human-computer interactive voice module will not alarm; when an object appears on the image plane and falls into the early warning alarm area, the human-computer interactive voice module will alarm and pass the TTS voice at the same time Broadcasting "Please pay attention to reversing" enhances the function of reminding the driver. At this time, the driver can also clearly know where the object is behind the car in the human-computer interactive display module, and the object is constantly flashing in the human-computer interactive display module, which is convenient for driving Member observation.
[0029] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, and substitutions can be made to these embodiments without departing from the principle and spirit of the present invention. And variations, the scope of the present invention is defined by the appended claims and their equivalents.
