Download Now Free registration required
The proposed system highlights a novel approach of detecting and tracking multiple objects in the cluttered area like crowd using greedy algorithm. The proposed framework uses position of traced low-level feature points to generate a group of autonomous rational mobility region as resultant. Various challenging factors towards the accuracy of detection rate for multiple objects are considered. The proposed approach has detected all the feasible rational mobile regions and extorts the sub-group which increases a total likelihood function along with assignment of each traced locus to one mobile region. Performance analysis is carried out with different set of video sequences to find that proposed system has gradual robust detection rate as well as highly cost-effective computationally.
- Format: PDF
- Size: 224.19 KB