An improved Gaussian Mixture Model with post-processing for multiple object detection in surveillance video analytics

Gaussian Mixture Model (GMM) is an effective method for extracting foreground objects from video sequences. However, GMM fails to detect the object in challenging scenarios like presence of shadow, occlusion, complex backgrounds, etc. To handle these challenges, intrinsic and extrinsic enhancement required traditional GMM. This paper presents a novel framework that combines improved with postprocessing multiple detection. In proposed system, parameter initialization considered improvement. Video preprocessing are improvements. Integration morphological operation helps better segmentation than GMM, it also increase detection performance by reducing false positives. process noise removal prepares input ready further processing. final step gradient operations used postprocessing. The approach was tested on surveillance sequences benchmark datasets such as PETS 2009 CD 2014(Change Detection). experimental results compared using ground truth evaluation metrics. show performs can effectively even illumination variation partial occlusion.