Method, terminal and system for correcting aberrant image

Abstract

The invention discloses a method for correcting an aberrant image, which comprises the following steps: obtaining an aberrant image at a random frame collected by a wide angle camera in the gesture variation process, and dividing the aberrant image at the random frame into image blocks with preset quantity; obtaining aberrant coordinates of target image blocks in the random-frame image; according to the aberrant coordinates of the target image blocks, looking up corrected display coordinates corresponding to the aberrant coordinates of the target image blocks from a generated mapping table which comprises the aberrant coordinate of each image block of the random-frame image and the corresponding corrected display coordinate; obtaining the same description parameter of the random-frame image and a former-frame image of the random-frame image; computing a first inter-frame relative depth of the random-frame image and the former-frame image according to the same description parameter; and outputting and storing the corrected display coordinates and the first inter-frame relative depth. The technical scheme provided by the invention can improve the precision of close range gesture identification.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a distorted image correction method, terminal and system. Background technique [0002] Among the many human-computer interaction methods, gestures, as an intuitive, natural, and easy-to-learn interaction method, have received extensive attention. Among them, the camera-based gesture recognition technology has received great attention due to its simple system structure and open application environment. It has won the favor and in-depth research of the majority of scientific researchers, and has achieved rapid progress and development in the industry. [0003] At present, camera-based gesture recognition technology has begun to be applied to the control operations of various scenes such as mobile phones and TVs, and plays an increasingly important role in the field of human-computer interaction. However, due to the relatively low control accuracy of the prior art...

AI Technical Summary

Problems solved by technology

Although gesture interaction based on dual cameras can solve the problem of close-range gesture recognition to a certain extent, because the stereoscopic viewing surface will become smaller when the dual cameras are close-range, the blind area of ​​stereoscopic imaging will become larger, and the depth information cannot be calculated well. , the accuracy of the parameters used to recognize gestures is still very low

Images