Improved Quantum Genetic Algorithm And Its Application In Image Matching

Quantum genetic algorithm is a kind of intelligent optimization algorithms thatquantum theory is applied to the genetic algorithm. It’s high parallelism and globalincreasingly is concerned, and is applied to a variety of optimization problems. Thisarticle is also based on the improved quantum genetic algorithm and its applicationto image matching.First, for the shortcomings of poor local search ability of thequantum genetic algorithm, this article raise hybrid algorithm of the triplecombination way by combining simulated annealing algorithm that a good localsearch ability with quantum genetic algorithm. simulation illustrates that the overallperformance of the three improved algorithm should be better than the standardquantum genetic algorithm, and one algorithm that simulated annealing algorithmapplied directly to part of the individual updated is the best. Then for the drawbackfor quantum genetic algorithm can not better deal with the optimization problem ofthe high dimension and many peaks, a new grouping of quantum genetic algorithmis rasied which the main idea is for every group records an optimal solution, and to acertain probability of exchange between the one-dimensional variable in theoptimal solution, and the use of unequal probability, improved step angle toimprove the overall performance of the algorithm. The experiments show that theimproved algorithm can effectively handle the high dimension and many peaksoptimization problems. The paper also proposed the normalized cross-correlationmeasure based on image gradient with the standard normalized cross-correlationmeasure compared with better anti-interference, easier to search to the correct matc hpoint. Finally, improved quantum genetic algorithm and the improved similaritymeasure is applied to the image matching and discussion image matching successrate, convergence speed, accuracy and other indicators. The results show that thequantum genetic algorithm as an image matching the search strategy is veryeffective and can greatly reduce the search time and improve search precision.