A forward intersection strategy on structural parameters for robotic vision system with 3 or more DOFs

The vision-based measurement for robot with 3 or more degrees of freedom (DOFs) require on-site calibrations and auxiliary sensors to establish an object coordinate system at present. Obviously, it is time consuming and requires extensive computing resources, and these hinder the application of vision-based measurement for robot in many fields. To solve these questions, a measurement coordinate system on robot was established according to the posture of robotic parts in preset position, structure of robot and vision system. Two categories of structural parameters were proposed; posture transformation matrices of robotic parts from the preset posture to an actual posture were built; and the visual-related spatial vectors were expressed by the two categories of structural parameters, optical parameters, and posture matrices in the measurement coordinate system. According to these, a new forward intersection strategy is proposed for robotic vision system. The new structural parameters for vision system on robot were proven to be reliable. The expressions of the visual-related vectors for robot were found to be correct and effective. Experimental results demonstrated that the new forward intersection strategy was accurate and efficient, and this strategy can be applied to robotic vision platforms to facilitate control of robots.