Abstract:Tridimensional attitude angle optical measurement is achieved by processing the optical image of the object"s surface to measure its angular posture. To improve the accuracy of tridimensional attitude angle optical detection, a design method for an optical measurement system based on stereoscopic vision for large field is proposed. The system uses a dual baseline parallel collimating light source and an attitude sensor, and functions such as spot segmentation and centroid positioning are achieved through components such as the image sensor and USB interface chip. On the basis of the Linux kernel, four software modules including signal collection, data porting, serial communication, and attitude angle analysis are designed, while stereoscopic vision algorithms are introduced to explore the correspondence between target points and image pixels, establish the tridimensional coordinates of spatial angles using the disparity theorem, and realize large-field tridimensional attitude angle measurement. Experimental results demonstrate that the system has small attitude angle calibration errors, strong time synchronization, and a measurement cumulative time of less than 15s, with high measurement accuracy and robustness, providing an effective technical means for measuring the precision parts of mechanical equipment.