Positioning platform based on face object position analysis

Abstract

The invention relates to a positioning platform based on the position analysis of human face objects, including: a regional positioning device, which is used to identify each human face area where each human face object is located in the airspace sharpening image based on human face imaging features, and is also used to identify each human face area based on the human face imaging feature Car imaging feature recognition of each car region where each car object is located in the airspace sharpened image; the centroid detection device is used to detect the difference between the centroid of the human face region and the centroid of the car region in the airspace sharpened image When the distance is less than or equal to the preset number of pixels, a signal of the existence of a retrograde vehicle is issued. The positioning platform based on the position analysis of the human face object of the present invention is simple in conception and convenient in operation. Due to the feature that the imaging area of ​​the front-row face object overlaps with the imaging area of ​​the vehicle object when the car is traveling in reverse, the full-color image capture action and subsequent image processing are performed in the direction that the car is allowed to drive in the direction of the one-way road, thus realizing the targeted recognition of retrograde behavior .

Description

technical field [0001] The invention relates to the field of object recognition, in particular to a positioning platform based on face object position analysis. Background technique [0002] Object recognition includes two tasks, classification and detection. Classification is used to judge whether an image contains certain types of objects, and detection requires marking the location and size of these objects. Object recognition is the key to understanding images and scenes, and has broad application prospects. It can be used in various fields such as automatic web image annotation, massive image search, image content filtering, robotics, security surveillance, and medical remote consultation. [0003] General object recognition faces many difficulties, and there is no perfect solution so far. These difficulties include: (1) Various factors such as illumination changes, viewpoint changes, scale changes, object deformation, occlusion, and background noise make the same objec...

AI Technical Summary

Problems solved by technology

[0003] General object recognition faces many difficulties, and there is no perfect solution so far. These difficulties include: (1) Various factors such as illumination changes, viewpoint changes, scale changes, object deformation, occlusion, and background noise make the same object appear in different images. (2) There are large differences between similar objects, which requires the recognition model to reflect the commonality between similar objects without confusing similar object categories; (3) A large number of categories increases the difficulty of system implementation