GPU based GMM segmentation of kinect data

TL;DR

This paper introduces a GPU-accelerated Gaussian Mixture Model method for real-time segmentation of RGBD data from Kinect, effectively handling illumination changes, shadows, and reflections with high accuracy.

Contribution

It presents a novel GPU-based GMM segmentation approach that fuses color and depth data for improved foreground detection in real-time.

Findings

Operates at 30fps, matching sensor frame rate

Robust against illumination changes, shadows, reflections

Effective fusion of color and depth streams

Abstract

This paper presents a novel approach for background/foreground segmentation of RGBD data with the Gaussian Mixture Models (GMM). We first start by the background subtraction from the colour and depth images separately. The foregrounds resulting from both streams are then fused for a more accurate detection. Our segmentation solution is implemented on the GPU. Thus, it works at the full frame rate of the sensor (30fps). Test results show its robustness against illumination change, shadows and reflections.