Detection of small moving objects based on motion vector processing using BDD method

Abstract

Motion Compensated Frame Interpolation addresses the tribulations of defective motion vectors that cause visual artifacts but cannot be detected by high residual energy or bidirectional prediction difference. A correlation-based motion vector processing method is proposed to detect those unreliable motion vectors by explicitly considering motion vectors correlation in the motion vector reliability classification, motion vector correction, and frame interpolation stages. MCFI can effectively discover the areas where no motion is reliable, such as occlusions and deformed scheme for the occlusion areas based on the analysis of their surrounding motion distribution. Interpolated Frame obtained using the proposed schemes have cleared structure edges and ghost artifacts which can be greatly reduced. MCFI shows better visual quality to obtain true motion. MCFI scheme is highly robust for the video sequences that contain fast motion.