Method and system for calculating head-to-body ratio generated by virtual image

Abstract

The invention relates to a method for calculating a head-to-body ratio generated by a virtual image. According to the method, a depth map and two color images (one is a low-definition ILD image and the other is a high-definition Ihd image, wherein the low-definition image is registered with the depth map) are obtained. The method for obtaining the ratio of the distance of the outer canthus of a human body to the height of the human body comprises the following steps of calibrating the focal length f of a high-definition camera in the direction of the double-eye connecting line; extracting thefeature point lld of a human face in the low-definition image; extracting the feature point lhd of the human face in the high-definition image; extracting the average depth ddepth of the two eyes in the depth map, so as to obtain the feature points lld of two eye positions; calculating the pixel euclidean distance dpixel between the feature points of two outer canthus in the human face in the high-definition image lhd; calculating the physical distance dphysics between the two outer canthus; and calculating the ratio alpha of the distance between the two outer canthus to the height of the human body. The method is fast in image acquisition speed and can quickly complete the human body five-sense measurement. In addition, during the image acquisition process, no radioactive substance is generated, so that the method is harmless to the human body. Meanwhile, the human body does not need to be in the naked state or worn in the approximately naked way. Therefore, the privacy of the human body is protected. The collection and the measurement do not need to be completed in a closed environment, so that the dimension space is saved. The method is beneficial to the mental states of measured people.

Description

technical field [0001] The invention relates to the technical fields of computer graphics and computer vision, in particular to a method and a system for estimating the ratio of the outer corner distance of a person to the height. Background technique [0002] With the development of information processing technology such as computer graphics, anthropometric measurement has attracted much attention as a research point and frontier direction in the field of computer graphics. Health, ancestry analysis (gene phenotype) and other aspects have been widely used and promoted. [0003] Currently, anthropometry can be divided into two categories according to the type of image: detection based on RGB images and detection based on depth images. The former has made great progress and is now the mainstream implementation method, but based on RGB images, it is easily disturbed by external factors such as illumination changes, shadows, object occlusions, and complex backgrounds, which gr...

AI Technical Summary

Problems solved by technology

The former has made great progress and is now the mainstream implementation method, but based on RGB images, it is easily disturbed by external factors such as illumination changes, shadows, object occlusions, and complex backgrounds, which greatly reduces the detection performance.

Images