Hand posture estimation method based on depth information and calibration method

Abstract

The invention discloses a hand posture estimation method based on depth information and a calibration method. The method comprises the steps that at the step S1, hand depth data is acquired, and a hand area is partitioned from the hand depth data; at the step S2, a palm posture is detected according to the hand area; at the step S3, positions of each articulation point of the hand are computed in combination with the palm posture and a hand standard bone model; at the step S4, projection characteristics of each articulation point of the hand are computed; and at the step S5, finger posture calibration is carried out according to the projection characteristics of each articulation point of the hand. The method disclosed by the invention is characterized in that the depth data is taken as the basis; the hand area is partitioned; the palm posture is computed; and the finger posture is estimated according to depth images through the posture calibration. The method is simple and practical.

Description

technical field [0001] The invention relates to the technical field of robot vision, in particular to a hand posture estimation method based on depth information and correction methods. Background technique [0002] In recent years, human-computer interaction technology has played an increasingly important role in life, and convenient and comfortable interaction methods can greatly enhance people's interactive experience. Although traditional interactive methods such as keyboard and mouse can meet daily interactive input, they are limited in terms of convenience and distance. Gesture technology has become a research hotspot in recent years, and there are endless ways to interact with hands. As the most flexible organ of the human body, the hand has a high degree of freedom and flexibility, so it can not only complete the work in daily life, but also meet the interactive input represented by gestures in the future. Among them, the gesture technology has also been developed ...

AI Technical Summary

Problems solved by technology

However, this method is based on the point cloud model established by the hand, and the three-dimensional gestures are varied and occluded, so it is difficult to complete the point cloud model, so this method is more complicated.

[0006] (2) A method and system for estimating three-dimensional gestures and attitudes based on depth data, the application number is 201510670919.6; the invention uses the Kinect sensor to obtain hand data, and on this basis, extracts features and adds a discriminant model to realize palm gestures and fingers respectively Pose regression, but the method is too complex

[0007] (3) A gesture recognition method and device, the application number is 201410036739.8; the invention obtains the color and depth data of the hand, uses the color image to complete the analysis of the hand depth data, and uses adaptive weighting to extract features in the hand contour area , based on the classifier to complete the type recognition of the current gesture, so as to recognize the gesture; this method is essentially an apparent method. For complex images in interactive scenes, it is not only difficult to solve the problem of complete segmentation of the hand area, but also unable to express Various gestures under high-dimensional hand posture

[0008] (4) Smart hand teleoperation control method based on Kinect human motion capture, the application number is 201410697456.8; this invention is also based on Kinect to estimate the hand posture, obtain hand depth data and color data, and use the color data to segment the hand area Finally, use the particle filter pose detection method based on template matching to predict the pose of the human hand, and then perform gesture recognition and interaction; this hand pose estimation method mainly realizes the estimation of the hand pose by tracking, and the real-time performance of the general algorithm is not good. And it is not good in the face of occlusion and large angle changes

Images