Research On Image Segmentation Algorithm Based On Fractional Order Differential

Image segmentation is the basic step of image processing and image recognition analysis.Therefore,how to obtain better image segmentation results and improve segmentation accuracy and efficiency has become one of the key issues in computer vision and digital image processing.In real life,images are inevitably affected by various factors such as noise and uneven illumination,resulting in phenomenon such as intensity inhomogeneous,blurred boundary,and weak texture in the image.The problem of intensity inhomogeneous,weak edge and weak texture image segmentation is still a difficult point in computer vision and medical image analysis.The segmentation algorithm based on the active contour model can better solve the problem of unsatisfactory results of such image segmentation.However,the active contour model has problems such as the selection of the initial contour of the evolution curve and noise sensitivity,which limits its application in practice.Therefore,this paper focuses on the problem of intensity inhomogeneous,weak edges,and weak texture image segmentation.The specific research work is as follows:1.The LIF model is sensitive to the selection of the initial contour of the evolution curve and the noise when segmenting intensity inhomogeneous,weak edges and weak texture images.To solve this problem,Image segmentation algorithm based on fractional differential and LIF model is proposed.A fractional differential gradient fitting term based on image global information is constructed,which is combined with the local intensity fitting term of the LIF model.The driving force of the curve evolution is composed of the global fractional differential gradient fitting force and the local intensity fitting force.Under the joint action of both,it moves to the target object boundary.It can be concluded from the experimental results that the improved model can handle intensity inhomogeneous,weak edges and weak texture images.The efficiency and accuracy of segmentation have also been significantly improved.At the same time,there is a certain degree of robustness to the initial position of the evolution curve and noise.2.Aiming at the CV model is not ideal for the segmentation of intensity inhomogeneous,weak edge and weak texture images,and the disadvantages of initial position selection of evolution curve and sensitivity to noise,A new CV combiningfractional differential and image local information is proposed.Firstly,the algorithm integrates fractional gradient information into the local information of the image,replaces the integer order global information of the CV model.Then an adaptive computational fractional order mathematical model is established based on the gradient modulus and the information entropy of the image.In order to avoid the reinitialization of the model,the model adds the constraint term of the symbol distance,thereby improving the evolution efficiency of the curve.Both the theoretical analysis and the experimental results show that the improved model solves the problem that the CV model is not ideal for the intensity inhomogeneous,weak edges and weak texture image segmentation,and it can adaptively determine the best fractional orders based on image features,the efficiency of the segmentation has also been improved.At the same time,it has certain robustness to noise and the initial contour position of the evolution curve.