Shape reconstruction system based on multi-view photogrammetry
Export citation
Abstract
Mexico has a growing need for lower limb prostheses due to a rising number of amputations caused primarily, by diabetic foot. Researchers enable functional and comfortable prostheses through prosthetic design by integrating new technologies applied to the traditional handcrafted method for prosthesis fabrication that is still current. That is why computer vision shows to be a promising tool for the integration of 3D reconstruction that may present to be useful for prosthetic design. This thesis has the objective to design, prototype and test of a functional system to scan plaster cast molds and serve as platform for future technologies for lower limb reconstruction applications. The image capture system is based on 5 stereoscopic RGB-D cameras with 3 DOF, as well as algorithms to calibrate the cameras to a global coordinate reference, to segment a target by removing the background, the conversion of RGB-D images into point clouds for later registration and surface reconstruction. The results for the plaster cast models show the segmentation and registration validations. The segmentation metrics of DC and HD show strong visual similarity with an average similarity percentage of 0.87 and average error of 6.40 mm respectively. Moving forward, the system was tested using a known 3D printed model obtained from a CT scan, to which comparison results via HD show an average error of less than 1.93 mm making the system competitive against the systems reviewed from the state-of-the-art.. The tests and results of the proposed system demonstrate it to be useful for the applications stated in the proposal, applications such as the generation of a digital library for plaster cast molds, serving as a platform for assisting the prosthesist during the manufacturing process, updating external geometries of previous CT scans to avoid radiation exposure and future extraction of viscoelastic properties of external tissue used for deformation analysis.