Multi-modal sensing positioning technology for intelligent verification of stacked kilogram weights
Di Zhao, Jian Ma, Wenbin Yu, Guixiong Liu · International Conference on Mechanical Engineering, Measurement Control, and Instrumentation · 2021
Unmanned verification of weights requires machines to grab and carry weights on behalf of people. High-precision spatial coordinates of weights are the key parameters for unmanned verification of weights. This article starts with multimodal sensing positioning technology, and proposes a spatial positioning technology solution based on multi-modal sensing stacking kilogram weights. Through technical solutions such as deep learning, LM optimization, and least squares fitting, the stacking of kilograms under fluorescent lighting is realized. The spatial positioning of the weight handle forms a versatile method for positioning the handles of stacked kilogram weights. Experiments were carried out by simulating the construction of 3×3×2 stacked kilogram weights in different states. The results showed that spatial positioning accuracy of the stacked weights instance is millimeter-level, and the calculation time of the overall algorithm is less than 6s, which can realize the spatial positioning of stacking kilogram weights in complex environments.