XZ-shape histogram for human-object interaction activity recognition based on Kinect-like depth image

Abstract

This paper introduces XZ-shape histogram in recognizing human performing activities of daily living (ADLs) which focuses on human-object interaction activities based on Kinect-like depth image. The evaluation framework was formulated in order to compare XZ-descriptor with previous shape histogram as well as X-shape histogram and Z-shape histogram. Each descriptor was segmented into several cases according to number of shells and symbols used in vector quantization process which was executed using our own dataset called RGBD-HOI. This study showed that XZ-shape histogram managed to outperform the other 3D shape descriptors along with the excellent one that compares the performance inferred by the area under receiver operating characteristic curve (AUC-ROC).The results of this study not only demonstrate the implementation of 3D shape descriptor in the dynamic of human activity recognition but also challenge the previous shape histograms in terms of providing low dimension descriptor that capable in improving the discrimination power of human-object interaction activity recognition.