Research On Multidimensional Image Segmentation Algorithm Based On Graph Theory

Image segmentation algorithm based on graph theory has perfect theoretical background and good segmentation effect,which has been paid more and more attention in image segmentation and computer vision.Traditional Graph Cuts segmentation algorithm may bring leakage segmentation and error segmentation.The segmentation effect of the algorithm needs to be improved,and the algorithm needs human interaction with low segmentation efficiency.In order to improve the shortcomings of traditional algorithms,the Graph Cuts algorithm of sequence image segmentation based on cosine similarity is proposed.The region term of the energy function is constructed using the maximum cosine similarity,and the maximum cosine similarity between the super-pixel and the seed clustering region is calculated.The boundary term of the energy function is constructed using the cosine similarity and color similarity,the color and relative distance feature similarity of neighborhood super-pixel is calculated.This algorithm is compared with the traditional Graph Cuts algorithm through the experimental verification,and the evaluation index shows that the accuracy and recall rate are improved,the false positive rate and the false negative rate is reduced obviously.Image segmentation algorithm based on Graph Cuts algorithm requires manual interaction,and it takes a lot of time and manpower.In order to improve the segmentation efficiency of multiple sequence images,this algorithm is applied to sequence image segmentation.The idea of the sequence image segmentation algorithm is to use the sequence image correlation to segment the sequence image segmentation,which is the segmentation result of the current image as the seed point set of the next image.In addition to the first image needs to be marked,the remaining images are saved for about 29 seconds,the more sequence images,the more time and manpower the algorithm saves.Traditional two-dimensional images can only show a section of the human body,but can not show the position in space,the shape,and the spatial relationship of the surrounding tissues and organs.Image segmentation technology towards multidimensional image segmentation development,not only in order to improve the accuracy of clinical diagnosis and the scientific nature of medical research,but also to reduce the storage of image data pressure.The three-dimensional reconstruction of the two-dimensional image after segmentation is carried out by using the surface rendering method and the volume rendering method respectively.The three-dimensional image is obtained and the multidimensional image segmentation is realized.And the system can be achieved rotation,drag,zoom in,zoom out and other functions,three-dimensional images can be viewed in all directions.