Sight angle estimation method based on human eye differential image

Abstract

The invention discloses a sight angle estimation method based on a human eye difference image, and the method comprises the steps: inputting two different human eye images: a tested human eye image and a reference human eye image into a twin neural network, and obtaining a difference value of a corresponding sight angle through network regression; when the sight angle of the reference human eye image is known, the sight angle of the tested human eye image can be obtained. The method solves the problem that in an existing sight line angle estimation method, the sight line angle is directly estimated from an image, and large estimation errors are caused by personal apparent differences. Meanwhile, the problem that the workload of data calibration for a certain individual is quite large is solved. According to the method, the difference value of the sight angles in the two images is estimated by using the twin neural network technology, so that the sight angle of the tested human eye image is obtained, the angle estimation precision is high, and the robustness is good.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to the design of a line-of-sight angle estimation method based on human eye differential images. Background technique [0002] Sight angle estimation is to calculate the line-of-sight angle of the human eye through image processing technology, which is generally divided into two estimation methods based on geometric model and appearance model. The estimation method based on the geometric model is to obtain the line of sight angle by calculating the geometric parameters of the eyes; the estimation method based on the appearance model is to regard this problem as a regression problem, and use machine learning technology to directly learn the regressor from the input image, so as to obtain line of sight angle. Since the former depends more on the resolution of eye images, the latter has become the mainstream research direction. However, in the current research bas...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to address the above-mentioned deficiencies in the prior art, and propose a line-of-sight angle estimation method based on the difference image of the human eye, so as to solve the problems of inaccurate estimated result of the line-of-sight angle of the existing human eye and heavy calibration workload

Images