The work conforms to the requirements of the times and aims to create a scar detection and surface area calculation software based on deep learning. The small program can process the image according to the scar processing picture of the patient taken by the doctor, transmit the picture to the server, use the TransUNet network to process the image, realize accurate segmentation of the scar edge, calculate the scar surface area, provide an objective basis for the doctor to charge according to the scar surface area, and also greatly improve the efficiency of scar treatment. At the same time, according to the detected scar edge and surface area and other information, the most appropriate scar drug dressing can be formulated for patients, which not only helps the treatment of patients, but also saves drugs to a certain extent. This software can be used in major hospitals, for all patients who need scar treatment, for medical staff in the treatment of scars. The scar cutting applet is divided into two applets: the doctor's side and the patient's side. Both applets can realize the registration, login, exit and addition of scar site labels for users. Scar pictures can be uploaded through photo albums or cameras to obtain the prediction results of scar area at the back end. The scar treatment effect can be observed through line charts of scar area changes on different dates. On the doctor side, doctors can see and manage their own treatment records of multiple patients. On the patient side, patients can choose their own doctors. After uploading information, the corresponding doctors can see and manage them. Between the corresponding doctors and patients, the treatment records can realize two-way management.