Motion Data Compression Research Articles

Feature Recognition of Human Motion Behavior Based on Depth Sequence Analysis

The current research on still image recognition has been very successful, but the study of action recognition for video classes is still a challenging topic. In this work, we propose a random projection-based human action recognition algorithm to address the lack of depth information in color information (RGB video frames) that is not easily affected by environmental factors such as illumination and the lack of ability to recognize actions along the direction of view. A network structure is designed to take the obvious advantage of long- and short-term memory networks for controlling and remembering long sequences of historical information. The network structure in this paper is constituted by multiple memory units. At the same time, this paper constructs the spatial features, temporal features, and depth features of the three recognition stream outputs into a feature matrix, whose feature matrix is divided into multiple temporal segments according to the temporal dimension, then inputs them into the network layer in order, and achieves the fusion of the feature matrix in this paper according to their correlation characteristics on the temporal axis. Here, we proposed the concept of random batch projection operators. This basically uses as much sublimitation information as possible to improve projection accuracy by randomly selecting several subdependencies as projections defined during projection. A compressed sensing design of human motion acceleration data for low-power body area networks is proposed, and the basic idea and implementation process of compressed sensing theory for human motion data compression and reconstruction in wireless body area networks are introduced in detail.

Computer animation in mobile phones using a motion capture database compressed by polynomial curve-fitting techniques

The aim of this paper is to efficiently implement motion-capture-based computer animation in mobile phones. Motion-capture-based computer animation is made from a large motion capture database stored in memory. Unfortunately mobile phones have little memory. To adapt motion-capture-based computer animation to the small memories of mobile phones, we revise a commercial motioncapture-based computer animation process to use a reduced motion capture database. To obtain a reduced motion capture database, we design a simple polynomial curve-fitting-based motion capture database reduction method. This approach preserves the smoothness of motion in the motion capture data, and does not overload the low computational capability of mobile phones. To further enhance performance, we optimize the degree of the polynomial fitting function. Using the developed approach we demonstrate an example implementation of motion-capture-based computer animation in a commercial mobile platform. Many motion data compression methods are already known, however this is the first attempt to specialize motion data compression for low performance architecture, mobile embedded system.

Wavelet Based Keyframe Extraction Method for Motion Capture Data

This research extracts keyframes from Motion Capture Data. Accurate selection of keyframes is a key point in motion synthesizing, editing, retargeting and motion data compression. Our method consists of two steps, curve selection and curve smoothing. In curve selection, every joint curve is re-represented with a simple numeric sequence based on wavelet transformation. Then the correlation between curves is calculated based on these numeric sequences. Several curves which are least correlated are chosen for keyframe extraction. In the next step, a noise filter is applied to selected curves. Finally the keyframes are selected from the de-noised curves.