Scholarlink

Every year, new requirements are imposed on modern industrial production, which relate to improving technological processes and product quality. A special place in new developments is occupied by research devoted to increasing the performance and durability of the cutting part of tools in automated production conditions. The authors have developed a new approach to improving the recognition quality of cutting part tool states with machine learning models. The article aims to create a new combined approach to improve the recognition quality of tool states with the machine learning models used. The scientific novelty consists of developing a segmentation method for automating the cutting part of tools segmentation that considers different features of cutting part tools with the machine learning models used. Practical usefulness consists in improving the quality of recognition of cutting part tool states with the machine learning models used. The vector-difference approach is used in the work to determine textural features. This approach makes it possible to obtain a specific texture feature - as a vector transformation of the features of different textures- based on vector algebra.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Combined Approach for Determining Tool Cutting Part States Using Machine Learning Models

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Oleksandr Derevianchenko,
  • Oleksandr Fomin,
  • Natalya Volkova,
  • Oleksiy Tataryn,
  • Isak Karabegovich

摘要

Every year, new requirements are imposed on modern industrial production, which relate to improving technological processes and product quality. A special place in new developments is occupied by research devoted to increasing the performance and durability of the cutting part of tools in automated production conditions. The authors have developed a new approach to improving the recognition quality of cutting part tool states with machine learning models. The article aims to create a new combined approach to improve the recognition quality of tool states with the machine learning models used. The scientific novelty consists of developing a segmentation method for automating the cutting part of tools segmentation that considers different features of cutting part tools with the machine learning models used. Practical usefulness consists in improving the quality of recognition of cutting part tool states with the machine learning models used. The vector-difference approach is used in the work to determine textural features. This approach makes it possible to obtain a specific texture feature - as a vector transformation of the features of different textures- based on vector algebra.