Human body part segmentation method for similar feature calculation of far infrared image

Abstract

The invention discloses a human body part segmentation method for similar feature calculation of a far infrared image. The method comprises the steps that a far infrared data acquisition module acquires human body far infrared images of multiple distances, multiple light rays and multiple scenes according to the characteristics of diseases; the data preprocessing module preprocesses the collected far infrared image; the region coordinate module proposes an Intete-UNet network model according to characteristics of human body region distribution, realizes segmentation of a human body region, and obtains a pre-training model of human body region segmentation through pre-training; the far infrared human body area pseudo RGB image module is used for acquiring pseudo RGB image areas corresponding to all areas of a human body; and identifying whether the human body sub-region pseudo RGB image in the far-infrared human body region pseudo RGB image module is abnormal or not through an abnormity identification module. According to the human body part segmentation method for similar feature calculation of the far infrared image, similar feature region segmentation is carried out on the human body region, and more effective temperature difference analysis can be carried out in combination with the distribution characteristics of the human body temperature.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to a human body part segmentation method for similar feature calculation of far-infrared images. Background technique [0002] Far-infrared imaging technology refers to the use of optical imaging principles to generate infrared radiation for detection targets. This technology is non-invasive, has no side effects, and can objectively represent the temperature of detection targets. Far-infrared images are collected by far-infrared imaging technology. The far-infrared radiant heat emitted by ) is converted into an intuitive temperature color map, so that the human body's functional metabolism can be recorded according to the temperature difference of different temperatures on the human body surface, and then the infrared heat radiation of normal tissue areas and abnormal tissue areas can be studied. Difference value, to judge the location and range of the lesion area o...

AI Technical Summary

Problems solved by technology

[0004] The main purpose of the present invention is to provide a human body part segmentation method for similar feature calculation of far-infrared images, which can segment similar feature areas for human body areas. Compared with conventional methods, it can combine the distribution characteristics of human body temperature for more effective Temperature difference analysis; in the convolutional layer of the classification sub-model, the lesion thermodynamic map corresponding to the feature map can be generated, and the area with obvious temperature difference difference can be visualized, which can help the user to have a certain understanding of the presentation of the temperature difference in the abnormal area, and then It can detect the initial state of the disease and suspicious lesion signals in time, and can intervene in time and carry out early and effective intervention on the disease, which can effectively solve the problems in the background technology

Images