An image non-target filtering method for ground plane identification recognition

Abstract

The invention discloses an image non-target filtering method for ground plane identification, and the method comprises the following steps: A, obtaining an inverse perspective image which comprises aground plane identification and a non-target; Processing the inverse perspective image IIPM by using an image processing technology to obtain a gray level image and a binary image; B, processing the gray level image by using a Canny edge detection algorithm to obtain an edge image; C, extracting the contour of each connected domain in the binary image to obtain a contour map; And step D, obtaininga contour image according to the edge image and the contour image. filtering out a non-target in the binary image; only the ground plane identification without height difference on the ground plane is reserved. According to the method, the final to-be-identified image which comprises the ground plane identification and filters out the non-target is obtained, when the image is used for ground plane identification identification, interference on later ground plane identification detection and identification can be well avoided, the identification rate of a later system on the ground plane identification is greatly improved, and the false detection rate is reduced.

Description

technical field [0001] The invention belongs to the field of image information processing, and in particular relates to an image non-target filtering method for ground plane mark recognition. Background technique [0002] In existing tasks such as road traffic sign recognition, parking space recognition, and 360-degree surround view systems, Metropolis first uses the inverse perspective transformation algorithm to transform the image captured by the camera to the ground plane, thereby obtaining a top view, and then in its binary value Detect and recognize signs such as zebra crossings, lane lines, guiding arrows, and parking spaces on the ground plane on the image. However, there are many interfering objects such as vehicles, pedestrians, and bushes that are not on the ground level in the real scene, which interfere with the recognition system, resulting in low final recognition rate and high false detection rate. [0003] However, the existing recognition technology based ...

AI Technical Summary

Problems solved by technology

However, there are many interfering objects such as vehicles, pedestrians, and bushes that are not on the ground level in the real scene, which interfere with the recognition system, resulting in problems such as low final recognition rate and high false detection rate.

[0003] However, the existing recognition technology based on inverse perspective image directly detects and recognizes the target in the image without considering filtering out those non-target interference areas, which leads to low recognition performance of the detection and recognition system

Images