Fast Adaptive Machine Vision Positioning Algorithm Based on Relative Threshold Features
Kai Zheng · 2022
With the vigorous development of machine vision, simultaneous positioning and methods based on vision have received extensive attention and research and are now commonly used positioning methods. In order to realize the problem of low efficiency and slow operation speed, a visual positioning algorithm is proposed. First, the algorithm uses relative thresholds for templates and sampled images, which effectively overcomes the impact and reduces the data. Secondly, the algorithm reduces the amount of calculation of the machine vision positioning algorithm, and adopts some methods to speed up and increase the speed, and can set some other conditions to further speed up. The algorithm can quickly locate the target in some ways, and then locate the center of the target faster, which could, to a large extent, accelerate machine vision and ensure the accuracy of the algorithm. The algorithm is fast and accurate, and can meet real-time requirements. Experimental research results show that pose estimation based on visual methods, compared with traditional pose estimation methods, uses relative threshold features as the main or only source of information, which can give a more detailed description of the environment in which the machine is located, and has good on-site performance characteristic.