出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：Chronic wound have a long recovery time, occur extensively, and are difficult to treat. Theycause not only great suffering to many patients but also bring enormous work burden tohospitals and doctors. Therefore, an automated chronic wound detection method can efficientlyassist doctors in diagnosis, or help patients with initial diagnosis, reduce the workload ofdoctors and the treatment costs of patients. In recent years, due to the rise of big data, machinelearning methods have been applied to Image Identification, and the accuracy of the result hassurpassed that of traditional methods. With the fully convolutional neural network proposed,image segmentation and target detection have also achieved excellent results. However, theaccuracy of chronic wound image segmentation and identification is low due to the limitation ofthe deep convolution neural network. To solve the above problem, we propose a post-processingmethod based on fully connected CRFs with multi-layer score maps. The experiment resultsshow that our method can be used to improve the accuracy of chronic wound imagesegmentation and identification.
