An automatic identification method of welding current and voltage based on visual sensing technology

In order to utilize the reliable welding current and voltage displayed on the welding machine and provide reliable input for the automatic control of welding current and voltage, it is necessary to study the automatic recognition of welding current and voltage displayed on the welding machine. Firstly, the image acquisition and preprocessing of welding current and voltage are studied, and the related research contents mainly include image acquisition, image grayscale, edge detection, morphological closing operation, and operation of removing small-area objects. Then, the character segmentation method of welding current and voltage is studied, and the main research contents include positioning display region, cropping display region, image binarization based on soft threshold filtering, image edge smoothing, and character segmentation method. Finally, the character recognition method of welding current and voltage is studied, and the main research content includes digit character recognition, decimal point character recognition, automatic recognition of welding current and voltage. The experimental results show that the proposed method can accurately identify the welding current and voltage values displayed on the welding machine, which is helpful to improve the efficiency and quality of automatic welding. A image containing a lot of environmental noise is used as the processing object, and it is found that the method researched in this paper can accurately identify the welding current and voltage value displayed on the welding machine, which is helpful to improve the efficiency and quality of automatic welding.