GLOBAL JOURNAL OF ENGINEERING SCIENCE AND RESEARCHES A HYBRID INTELLIGENT MODEL FOR CROWD DENSITY ESTIMATION

Ali Salem, Ali Bin Sama, Hussein Salem, Ali Bin Sama
2017 unpublished
This work is aiming at the development of a hybrid intelligent model to tackle the problem of crowd density estimation. The presented hybrid model comprises Extreme Learning Machine (ELM) for pattern recognition, Differential Evolution (DE) for model construction, as well as texture feature extraction techniques for input pattern encoding. In this work, DE is adopted to design an efficient recognition model by performing training instances selection as well as ELM topology selection. To assess
more » ... lection. To assess the performances of the proposed model, a three popular crowd density benchmark dataset are used in this study including PETS_2009,Chunxi_Road, and Mall dataset.
fatcat:bgq5oyg52zdvjphbaacpuzyzcq