A Free Viewpoint Video Playback Method Based on Somatosensory Camera Gesture Recognition

Abstract

The invention relates to a free-view video playback method based on gesture recognition of a somatosensory camera, which acquires a frame of depth image and human skeleton data, preprocesses the depth image with a bilateral filtering method, calculates the center position of the right palm and the position of the right arm, and obtains the right palm. Contour, extract the contour feature of the right palm and the position feature of the right arm to judge the gesture, and convert the recognized gesture into the corresponding viewpoint playback operation. The invention uses the bilateral filtering method to preprocess the depth image, so that the generated depth image better retains the edge contour information in the image. A simple and fast gesture recognition method is designed, which effectively reduces the jamming of user gestures when switching viewpoints. A new method for zooming in and zooming out and switching between viewpoints is designed. By recognizing the user's gestures, the part of the viewpoint image is zoomed in and out and the viewpoint is switched, which can bring users a good integration when watching free viewpoint videos. sense.

Description

technical field [0001] The invention relates to the field of free-viewpoint video playback, in particular to a free-viewpoint video playback method based on somatosensory camera gesture recognition. Background technique [0002] With the development of information technology, people have more and more stringent requirements for video playback, from the initial black-and-white video image to the current color video image, from the initial SD and HD to the current ultra-definition and ultra-high-definition, but people There is a physiological limit to the vision of the eye, and the demand for resolution and chromaticity is about to become saturated. In addition, the two major development directions of video and images in the 21st century are high-definition and three-dimensional. High-definition has been basically realized at present, and three-dimensional will be the development direction of video and images in the future. The display methods of stereoscopic video can be div...

AI Technical Summary

Problems solved by technology

The steps of this gesture recognition method are relatively complicated, the calculation complexity is high, and it is not conducive to the real-time switching of free viewpoints.

Images