Human Motion Detection Using Fuzzy Rule-base Classification Of Moving Blob Regions

The task of detecting and classifying human motion is an important preliminary tool for many high-level applications. However, many approaches suffer from the lack of robust classification and proper motion cues. This paper presents a novel human motion detection algorithm that uses a fuzzy rule-base classification scheme based on moving blob regions. This approach first obtains a motion image through the acquisition and segmentation of video sequences. Then, preprocessing is applied to the motion image before major blobs are identified. Using motion estimation and ellipse fitting, three blob characteristics are extracted from the major blobs as classification criteria. These characteristics are used as inputs to a fuzzy rule-base for classification of the detected motion. Through experimental evaluation – a database test and real-time field test, the implemented system achieved good detection rates in both tests at efficient real-time speeds. In comparison with earlier approaches, this algorithm also managed better detection rates. Above all, the performance of the proposed algorithm has demonstrated its feasibility for an effective real-time implementation.