贫液电池整理装置的视觉检测系统研究

Research on Visual Inspection System of Lean Battery Sorting Device

根据废旧贫液电池的智能化拆解工艺要求,需要将贫液电池整理为稳定待切割状态,为实现该目标,设计了一种基于机器视觉的整理上线装置。该装置首先通过视觉相机采集到贫液电池的图像,然后运用Sobel算子对图像进行卷积运算得到梯度图,采取梯度投影法求出边缘位置并结合最小外接矩形、霍夫变换结合滤除算法对预处理后的图像分别实现尺寸测量和极柱面识别,进而完成对贫液电池状态的判断,最后利用三向旋转部件将贫液电池调整为稳定待切状态。经实验验证,平均每次视觉检测的时间在1s内,尺寸检测误差值在1mm内,满足该设计在时间和精确性上的要求,有效减轻了人工劳动强度、提高贫液电池回收效率。

According to the requirements of the intelligent dismantling process of used lean-liquid batteries,it is necessary to organize the lean-liquid batteries into a stable state to be cut.To achieve this goal,a machine vision-based sorting and online device is designed.The device first collects the image of the lean battery through the vision camera,and then uses the Sobel operator to convolve the image to obtain the gradient map.The gradient projection method is adopted to find the edge position and combined with the smallest bounding rectangle.Hough transform and filtering algorithm Realize size measurement and pole surface recognition on the pre-processed image,and then completes the judgment of the state of the lean battery,and finally adjusts the lean battery to a stable standby state with a three-way rotating part.Experiments have verified that the average visual inspection time is within 1s,and the size inspection error is within 1mm,which meets the requirements of the design in terms of time and accuracy,effectively reduces the labor intensity and improves the efficiency of lean battery recovery.