multi-objective optimization of vision metrology camera placement based on pareto front concept by nsga-ii method

nowadays, the subject of vision metrology network design is local enhancement of the existing network. in the other words, it has changed from first to third order design concept. to improve the network, locally, some new camera stations should be added to the network in drawback areas. the accuracy of weak points is enhanced by the new images, if the related vision constraints are satisfied simultaneously. therefore, the camera placement is an optimization problem that here is solved by using nsga-ii, a multi-objective evolutionary algorithm (moea) based on pareto front concept. although we have proposed two deterministic ito and oti methods and a non-deterministic fuzzy camera placement method in our previous research, here we solved the problem by an moea method. the nsga-ii network design method is able to solve the problem in complex cases in which other aforementioned methods are failed or cannot converge to global optimum. in addition, it is a good means to analysis the capabilities of other methods especially in complex network cases. it also gives us several optimal solutions for camera placement, so that designer can select one of them based on his/her experience and environmental restrictions. in this research, we did various tests on a complex example of camera placement by using nsga-ii algorithm. the result demonstrates the high capabilities of the method in solving and analyzing the camera placement in complex close-range photogrammetric networks.