This research presents a "Multi-camera for pilot's cockpit measurement system", which uses four multi-view images to eliminate the instrument and human body shielding and record the touched area. That could record the body reaction time (velocity and acceleration) and trajectory of the tested personnel. Real-time conversion of multi-view images corresponding to the 3D skeletal joint coordinate information of the human body, which measure the human-computer interaction human factors engineering integration of limb reaction time and trajectory measurement system. Finally, make prototypes, test and optimize, and achieve the research on the optimal cockpit touch area by conducting multi-view image simulation feasibility experiment framework and measurement process method. Using multiple depthsensing cameras to perform low-cost, standardized automatic labeling of human skeleton joint dynamic capture.