泵机金属外壳的缺陷检测数据

本数据基于特定的泵机金属外壳的缺陷检测方法获得的缺陷数据集。包括获取泵机金属外壳灰度图；获取每个像素块的梯度显著性；获取每个像素块的灰度显著性；根据每个像素块的灰度显著性和梯度显著性获取每个像素块的综合显著性；根据每个像素点增强后的灰度值获取增强后的图像；根据检测出每个增强后边缘线中像素点的数量判断是否存在缺陷并生成缺陷数据集。

This defect dataset is derived from a specific defect detection method targeting the metal housings of pumps. The workflow includes: acquiring grayscale images of pump metal housings; extracting gradient saliency for each pixel block; extracting grayscale saliency for each pixel block; calculating the comprehensive saliency of each pixel block based on its grayscale and gradient saliencies; generating enhanced images using the enhanced grayscale values of individual pixels; and determining the presence of defects and generating the final defect dataset by counting the number of pixels within each detected enhanced edge line.

搜集汇总

数据集介绍

特点

该数据集是由济宁安泰矿山设备制造有限公司提供的泵机金属外壳缺陷检测数据，基于灰度图和像素块分析生成，主要用于球墨铸铁领域的表层缩孔和缩松缺陷检测。