Mapping Classification Results on 3D model: a Solution for Measuring the Real Areas Covered by Skin Wound Tissues

we present in this paper part of our work inside the ESCALE project dedicated to the design of a complete 3D and color wound assessment tool using a simple hand held digital camera. The computation of a 3D model for wound measurements using uncalibrated vision techniques has already been investigated in the project. This article presents our methodology to classify wound tissues in the color images, in order to combine shape, deduced from the 3D model, and color analysis in a single tool for real tissue surface measurements. As a first step, we have adopted an original approach based on unsupervised segmentation prior to classification, to improve the robustness of the labelling stage. In the second step, the tissue classification is directly mapped on the mesh surface of the wound to measure real tissue growth and changes. For this purpose, a database of different tissue types has been built and the ground truth of the images is provided by the fusion of several clinician manual labellings. Then, color and texture tissue descriptors are extracted from tissue regions of the images database, for the learning stage of an SVM region classifier relying on expert ground truth. The SVM prediction model has been used to label the segmented regions of the database. Finally, we map the 2D tissue classification on the 3D model. Clinical tests demonstrate that the monitoring of the healing process is very accurate compared to single view analysis.