METHOD FOR OPTIMIZATION OF VIDEO COVERAGE

摘要

Optimizes the positions and angular orientations of cameras (102) used to cover a predetermined volume (100) such as a hall or sports field by combining a genetic algorithm and a simulated annealing algorithm. First, random initial solutions are generated and a local search is performed around each solution to find a local optimum solution. Then each local optimum solution is mutated randomly. The mutated solutions are combined and sorted by coverage level. The mutated solutions having the highest coverage levels are selected for simulated annealing. The simulated annealing algorithm generates a solution within a predetermined search radius of the mutated solution. A coverage level is calculated for the new solution. The simulated annealing algorithm repeats until global optimization is achieved. Each solution at each state is represented by a matrix whose orders equal the number of cameras (102) to be placed and the five degrees of freedom of each camera.