Mixed reality depth contour occlusion using binocular similarity matching and three-dimensional contour optimisation

Abstract Mixed reality applications often require virtual objects that are partly occluded by real objects. However, previous research and commercial products have limitations in terms of performance efficiency. To address these challenges, we propose a novel depth contour occlusion (DCO) algorithm. The proposed method is based on the sensitivity binocular stereoscopic vision device. In this method, map combined with sparse obtained from two-stage adaptive filter area stereo matching algorithm extracted digital image stabilisation optical flow method. We also quadratic optimisation model three constraints to generate an accurate dense for high-quality real-virtual occlusion. whole process accelerated GPU. evaluate effectiveness algorithm, demonstrate time consumption statistical analysis each stage DCO execution. verify reliability effect, conduct experimental single-sided, enclosed, complex occlusions. Subsequently, compare it without optimisation. With our GPU implementation real-time DCO, evaluation indicates applying presented enhances visual quality