| At present, the application of infrared imaging in medicine has been very broad, but some research about infrared image and the lesion of knee aren't widely. Especially the infrared image analysis and research through the multi-scale entropy algorithm is still at the international forefront of research areas. In this paper, we will focus primarily on exploring the application of multi-scale entropy about infrared image analysis at the basis of infrared images research. Based on the multi-scale entropy, the study indicates the multi-scale entropy have more advantages than other entropy algorithm on physiological signal processing. Through analyzing the multi-scale entropy theory, the paper proposes the implementation steps about the multi-scale entropy algorithm. The multi-scale entropy algorithm analyzes the pixels sequence about image in multiple scales. It tries to illustrate the change rule among images through complexity of pixels sequence. This article demonstrates comparative study about the infrared images contrast between pathological knee and reference group through the multi-scale entropy algorithm, and at last it will find the rule of variation between infrared image and pathological changes of level about knee lesion in patients. The research results provide the foundation for the clinical diagnosis and also discuss the value of application about the multi-scale entropy algorithm in infrared image analysis.This research regards the knee as the object of study, and discusses infrared image analysis problem to the disease area of knee joint of human. It mainly explores from the content blow:It is the work of image collection about infrared thermal. Based on the thermal infrared imager selection, the model of thermal infrared imager was chose. In the process of image acquisition, the infrared image will achieve the best results through performance correction of thermal infrared imager.It provides a new kind of method which can process the Infrared image through the multi-scale entropy algorithm. It provides a new kind of method which can process the Infrared image through the information entropy algorithm. Because of the lesion, the surface temperature of knee-joint is different. This research uses the multi-scale entropy algorithm for processing the pixel sequence of infrared image, then it obtains the complexity of sequences, we can judge the degree of knee lesions through comparison of complexity. Based on the analysis of infrared image about the knee joint lesions, the multi-scale entropy algorithm can be applied on analysis of the knee joint infrared image, and the paper finds the standard multi-scale entropy value curve about the normal knee infrared image. When the multi-scale entropy value curve of infrared image of knee are less than the standard multi-scale entropy value curve in the most scales, it shows the knee-joint may exist pathological changes, and the entropy is smaller than the standard entropy, the pathological changes are serious. |
