刀具损伤图像数据集

研制出一套具有镜头保护与清洁功能的刀具损伤图像在位采集装置，将视觉系统保护装置，刀具清洗系统与数控机床集成，实现刀具图像在位采集，避免拆卸刀具离线监测，兼顾检测精度与效率。基于已搭建的基于机器视觉的刀具损伤在位监测系统，针对立铣刀开发了基于大数据的刀具损伤图像数据集。目前数据集已有超过50张原始图片，数据集中刀具损伤类型覆盖磨损、破损、崩刃、裂纹等常见类型。数据量为5.87MB。

A dedicated in-situ collection device for tool damage images with lens protection and cleaning functions was developed. The vision system protection device, tool cleaning system and CNC machine tool were integrated to achieve in-situ collection of tool images, avoiding offline monitoring by disassembling the tool while balancing detection accuracy and efficiency. Based on the established machine vision-based in-situ tool damage monitoring system, a big data-based tool damage image dataset was developed for end mills. Currently, the dataset contains more than 50 original images, covering common tool damage types such as wear, chipping, edge breakage, cracks and other typical damage patterns. The total size of the dataset is 5.87 MB.

搜集汇总

背景与挑战

背景概述

该数据集是一个专注于刀具损伤的图像集合，基于机器视觉在位监测系统开发，针对立铣刀，包含超过50张原始图片，覆盖磨损、破损、崩刃、裂纹等常见损伤类型，数据量约5.9MB。它适用于机器视觉和深度学习领域的损伤检测研究，来源于国家重点研发计划项目，旨在支持切削加工工艺的优化和预测维护。