Date of Graduation
Statler College of Engineering and Mineral Resources
Lane Department of Computer Science and Electrical Engineering
Continuous monitoring of changes in wound size, wound area, and volume, is key to predict whether wounds will heal on time. Wound measurement methods can be subdivided into non- contact and contact methods. Contact methods are prone to errors given the human intervention and it increases the chance of discomfort during measurement. Alternative methods, such as image- based non-contact methods, eliminate any discomfort and have good reliability for measuring a wound. However, existing image-based non-contact methods are expensive. This is because these methods build a 3D model of the wound using expensive devices in order to allow the clinician to obtain the necessary wound measurements. To alleviate the cost of these systems, the proposed system described in this report measures wounds using low-cost depth cameras such as the Microsoft Kinect. This report describes methods that take in an RGB image from the Microsoft Kinect, computes the necessary parts of a 3D wound model, and finally reports wound measurements. The proposed system requires the user to draw the contour of the wound on the image. Then, the system automatically extracts all the necessary information from the RGB and depth images to create a minimal 3D model of the wound. Subsequently, the system processes the 3D model in order to facilitate the estimation of the wound area and volume. Finally, the system reports the measurements to the user. This report presents experiments demonstrating that the proposed system achieves acceptable measurements despite the fact that it uses a low-cost and noisy imaging sensor.
Gao, Xiang, "Efficient wound assessment system with an RGB-D camera" (2018). Graduate Theses, Dissertations, and Problem Reports. 3989.
Available for download on Friday, January 01, 2021