ADVANCED MOTION DETECTION APPROACHES OF REAL-TIME VIDEO PROCESSING USING GAUSSIAN AND Σ-∆ MODELS
M.M. Abutaleb · INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING & TECHNOLOGY · 2025
Motion detection in video sequences is a critical task in computer vision, with applications ranging from surveillance to autonomous systems.This paper presents two efficient algorithms for motion detection: the Gaussian Single-Pixel Multiple-Distributions (SPMD) algorithm and the Σ-∆ SPMD algorithm.Both methods aim to reduce computational complexity while maintaining high accuracy in detecting moving objects.The Gaussian SPMD algorithm models each pixel as a mixture of Gaussian distributions, while the Σ-∆ SPMD algorithm employs Σ-∆ modulation for background estimation.Experimental results demonstrate that the proposed algorithms significantly reduce processing time compared to traditional methods, such as the Stauffer algorithm, while maintaining competitive accuracy.The Gaussian SPMD algorithm achieves a 42% reduction in processing time, and the Σ-∆ SPMD algorithm further improves frame rates, making it suitable for real-time applications.This study provides a comprehensive