Human Face Localization in 3D For Humanoid Robot Vision

Reliable three-dimensional localization of human locations is critical in a variety of computer vision applications, including robots, self-driving automobiles, security, and human-computer interaction. This research focuses on improving the vision system of Arslan humanoid robot and allows it to maintain a constant connection with people in its field of view. The system uses human binocular vision principles to extract depth information from photos taken with two pan cameras calibrated with servo motors for accurate angle alignment, as well as two identical USB cameras. The methodology consists of taking photos, recognizing faces, comparing face counts amongst cameras, and conducting 3D localization. The basic goals are to keep eye contact with individuals, estimate their approximate locations, and correctly establish their precise places. Methodological breakthroughs have greatly increased localization accuracy, lowering mistakes and increasing dependability. These advancements open the door for more responsive and intuitive human-computer interactions, benefiting the larger area of computer vision technology