Improvement Of Intelligent Optimization Algorithm And Its Application In Image Segmentation

Image segmentation is the process of dividing an image into several non-overlapping sub-regions and extracting meaningful regions.In the field of digital image processing and computer vision,image segmentation is the key step from image processing to image feature extraction and recognition.Its quality affects the subsequent image analysis and pattern recognition.Therefore,fast and efficient segmentation methods have always been a research hotspot that researchers pay close attention to all the time.Threshold segmentation is an effective and practical segmentation method.Its principle is to find the optimal threshold according to some criteria,and then segment the image according to the obtained threshold.In the practical application process,in order to meet our needs,sometimes we need to use multi-dimensional threshold segmentation method or select multiple thresholds at the same time to achieve effective segmentation of the target image.With the increase of image information dimension or the number of selected thresholds,the computational complexity of threshold segmentation algorithm increases rapidly,which greatly improves the computational time and limits the application scope of the algorithm to a certain extent.For this reason,this paper proposes two improved intelligent optimization algorithms and the improved algorithms are used in image threshold segmentation.Paper research content is as follows:In order to overcome the problems of low segmentation accuracy and poor anti-noise performance of basic artificial bee colony algorithm in noisy image segmentation,a two-dimensional Otsu image segmentation was proposed based on improved artificial bee colony algorithm.First of all,this algorithm improves the convergence speed and the global search ability by introducing the global best position to guide the search direction and proposing different search strategies in the employed bees phase and onlooker bees phase.Secondly,to prevent bees from falling into local best during the searching process,simulated annealing mechanism was introduced into the artificial bee colony algorithm and the Metropolis rule was used to update nectar source position.Finally,2-D Otsu threshold segmentation method combined with improved artificial bee colony algorithm was applied to searching the optimal threshold.In order to test the performance of the algorithm,several different kinds of gray-scale images and noisy images were selected for simulation experiments,and the results were compared with those of exhaustive search,artificial bee colony algorithm and firefly algorithm.Experimental results show that this algorithm overcomes the time-consuming shortcomings of exhaustive search method,and the segmentation effect is better than artificial bee colony algorithm and firefly algorithm.It has better anti-noise performance and can effectively solve the real-time segmentation problem of noisy images.In order to overcome the issue of large amount of calculation and large computing time of the traditional multi-threshold image segmentation methods,A multi-threshold image segmentation method based on improved cuckoo search algorithm was proposed.this algorithm improves cuckoo search algorithm by teaching-learning search strategy,elite adaptive competition sharing mechanism and simulated annealing mechanism.Then maximum entropy multi-threshold segmentation method combined with improved cuckoo search algorithm was applied to searching the optimal threshold.In order to verify the effectiveness of the improved method,several different types of complex multi-target images are selected for segmentation experiments,and compared with four other advanced segmentation algorithms.Experimental results show that the proposed method is superior to the contrasted algorithms in segmentation accuracy,running time and convergence.It can quickly and effectively solve the multi-threshold segmentation problem of complex multi-target images.