Research Of Medical Image Superpixel Segmentation Algorithm Based On Std_SLIC

With the rapid development of medical imaging technology,medical imagesplays an increasingly important role in scientific research and medical diagnosis and treatment.As an important means of dealing with medical images,medical image segmentation technology plays an important role in the field of medical image processing and has drawn the widespread attention of experts and scholars in related fields.As a collection of pixels with similar characteristics in an image,the number of superpixels is much smaller than the number of pixels.For a medical image with more and more pixels,using superpixels instead of pixels as image primitives can greatly improve the image processing speed,and superpixels can also reflect part of the space relationship between the pixels in the image and the characteristics of the relationship can improve the accuracy of the target extraction.In this paper,we mainly study and improve the application of SLIC(super linearly interpolated clustering)algorithm in superpixel segmentation of medical images.This paper aims at the accuracy of medical image for superpixel segmentation results and improves the updating strategy of traditional SLIC superpixel segmentation algorithm in iterative updating of cluster centers.When the cluster center is updated,only a pixel cluster with a similar gray level to the original cluster center is used to calculate a new cluster center,and a std_SLIC super pixel segmentation algorithm for the medical image is proposed.The method firstly uses the three-dimensional histogram denoising model to process the original medical image and reduces the noise interference in the image.Then the gamma enhanced model is used to increase the contrast between the target area and the background area,and the probability of dividing pixel errors at the boundary of the target area is reduced.Finally,the superpixel segmentation of the medical image is performed with the std_SLIC algorithm which improves the updating strategy of the cluster center,and the final superpixel segmentation result is obtained.This method reduces the influence of pixels with dissimilar features on the process of calculating a new clustering center and makes the clustering center more accurate.Experimental results show that the proposed algorithm has a significant improvement in the accuracy of medical image superpixel segmentation.Like the SLIC algorithm,the std_SLIC algorithm uses pixel color information and spatial locations to calculate the similarity between pixels.Most medical images are grayscale images.The color information that can be applied is only the gray value of the pixel.However,different tissues and organs have their own unique texture features.These texture features can distinguish different tissues and organs,and the change of the texture features can also be done reflect the characteristics of human organs and pathological changes,the medical research,medical diagnosis and clinical treatment are of great help.Therefore,applying the texture features of an image to the calculation of the similarity between pixels can greatly help improve the accuracy and accuracy of the superpixel segmentation of medical images.Due to the complexity of the human body structure,the texture features in medical images are very complex.Therefore,this paper uses the local texture feature extraction method to extract the texture features in medical images.After comprehensively analyzing some local texture feature extraction methods,the Ltridp algorithm's amplitude model is selected to extract the texture features of the image.After denoising and enhancing the medical image,the extracted texture feature information is added to the std_SLIC superpixel segmentation algorithm to calculate the similarity between pixels in the process,improve the accuracy of superpixel segmentation.The experimental results show that the std_SLIC superpixel segmentation algorithm which combines the texture features of the image has further improved the accuracy of segmentation.