CROWD DENSITY ESTIMATION USING ACCUMULATED MOSAIC IMAGE DIFFERENCE

We present a bidirectional people counting system based on computer vision and propose solutions to various common problems such as the occlusion of people and discriminating between people and objects such as shopping trolleys or bags in stores. We propose the use of the extended modified condensation algorithm, based on optical flow generated from the movement of people and depth to the height of the system, as an estimation method for multiple people. The inclusion of different features relevant to people tracking, such as movement, size, and height, adapting the propagation and observation models in the particle filter and followed by a clustering method, provides sufficient accuracy and robustness to achieve high counting rates. It reduces communication cost and Data Retrieval is easy.