Eyeglass detection method and device

Abstract

The invention provides an eyeglass detection method and device. The method concretely comprises the following steps of: acquiring an eyeglass region image from a face image to be detected; carrying out primary discrimination on the eyeglass region image according to a single-eye region classifier, and detecting according to a primary discrimination result, wherein the detecting process comprises the following steps of: if the absolute value of the primary discrimination result is more than a preset confidence threshold, outputting a detection result indicating whether a pair of eyeglasses are worn or not according to the primary discrimination result; and if the absolute value of the primary discrimination result is less than or equal to the preset confidence threshold, carrying out secondary discrimination on the eyeglass region image according a double-eye region classifier, and outputting the detection result indicating whether the eyeglasses are worn or not according to the secondary discrimination result. By adopting the method provided by the invention, the requirement on an initial face position is low, and the accuracy of eyeglass detection can be improved.

Description

Technical field [0001] The present invention relates to the technical field of pattern recognition, in particular to a method and device for detecting glasses. Background technique [0002] Glasses are the most common accessory. It detects whether glasses are worn in the face image, and can calculate the position and shape of the glasses, which is important for improving the performance of face recognition algorithms, face synthesis, face cartoons and face recognition Practical value. For example, in face recognition, if the position of the glasses is detected in a face wearing glasses, a naked face without glasses can be obtained by removing the glasses. [0003] Due to the diversity of glasses shape and image quality, there are many difficulties in glasses detection. For example, early glasses detection mainly used image processing and template matching methods to detect the lower frame of the glasses and the bridge of the nose according to the discontinuous changes in the pixe...

AI Technical Summary

Problems solved by technology

However, due to the influence of different light, the image obtained by image processing and Hough method after imaging is excessively dependent on the edge of the image, so there is noise, and noise interference will often make it impossible to obtain feature points or accurate feature points, so the accuracy of detection relatively low

[0004] In order to improve the accuracy of detection, some existing methods try to use wavelet features to detect glasses, but wavelet transform is a localized analysis of time and frequency. time subdivision at low frequency, and frequency subdivision at low frequency; according to the principle that "a small loss can make a thousand miles away", if the initial face position is inaccurate, it will affect the accuracy of information on various scales and directions. Face position requirements are very high

Images