Research On Image Segmentation Method Based On Superpixel And Its Application

In digital information age,clustering is a common method in image segmentation.Clustering can divide the data into non-coincident clusters by analyzing the internal relations of the data,it can divide the image information into non-coincident parts to complete the image segmentation.FCM algorithm has a good allocation of overlapping data,and can get better boundary segmentation results for image segmentation.However,the FCM algorithm is sensitive to noise and lacks full utilization of spatial information,so the results on some images are not ideal.The introduction of superpixel segmentation algorithm can improve the segmentation performance.This paper proposes an image segmentation algorithm based on superpixel segmentation,and studies how to get better segmentation results.(1)An image segmentation method SLICHT-FCM-PCM is proposed in this paper,which combines the simple linear iterative clustering(SLIC)algorithm with multiple features and the algorithm combined by FCM and PCM(FCM-PCM).Firstly,the local homogeneity features and texture features are fused into the features of the traditional SLIC algorithm,and an SLIC superpixel segmentation algorithm(SLICHT)that integrates multiple features is proposed.Then,the superpixel blocks obtained by SLICHT superpixel segmentation algorithm are clustered and merged by FCM-PCM algorithm to achieve image segmentation.(2)Aiming at the problem that SLICHT-FCM-PCM algorithm involves multiple parameters and the parameters have great influence on the segmentation results,this paper proposes an improved grasshopper optimization algorithm(WGOA),which uses the WGOA algorithm to optimize the SLICHT-FCM-PCM algorithm to obtain better segmentation results.(3)The validity test was carried out using the MSRC dataset and ECSSD dataset,and the accuracy and F1 score of the segmentation results were well performed.This paper also applies the algorithm to practical problems to segment the self-built fabric images,and the algorithm has good performance in segmentation accuracy and visual effect.This paper proposes an image segmentation method based on superpixel.The method uses the SLIC algorithm with multi-feature fusion for superpixel segmentation,and the FCM-PCM algorithm is applied to image segmentation for the first time.The improved locust optimization algorithm is also used to optimize the feature weight parameters to achieve adaptive segmentation.