Scholarlink

<p>To address the challenges associated with capturing microdamage images of silicon nitride bearings, particularly those characterized by blurred edges and distorted contours established an experimental platform is established for acquiring such microdamage images. Furthermore, a comprehensive segmentation approach is developed, tailored to the microdamage features of silicon nitride bearings. In this process, the sigma function is optimized to eliminate the influence of zero-value elements, effectively reducing image noise and improving overall image quality. Additionally, iterative refinement of contour lines is implemented to ensure accurate delineation of actual microdamage boundaries, thereby enabling precise segmentation of microdamage images. The images processed with dynamic scale median filtering and the active contour iteration model exhibit a peak signal-to-noise ratio of 44.66 dB, an average structural similarity of up to 97.63 %, an average extraction accuracy of 98.07 %, and an average edge overlap of 93.68 %, effectively reducing the impact of artifact contours and edge collapse features on the extraction of damaged images. This method has improved the efficiency of failure analysis for silicon nitride bearings, provided a safeguard for the development of image processing technology, ensured the safe use of silicon nitride bearings in the field of new energy vehicles, and indirectly promoted the development of the new energy sector.</p>

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

Microdamage extraction method of silicon nitride bearing based on image processing technology

  • 期刊论文

  • 发表日期：

    2025年12月4日

查看全文
  • Qi Zheng,
  • Miao Mei,
  • Xiaobin Cui,
  • Jiao Li,
  • Weiwen Hu,
  • Nanxing Wu

摘要

To address the challenges associated with capturing microdamage images of silicon nitride bearings, particularly those characterized by blurred edges and distorted contours established an experimental platform is established for acquiring such microdamage images. Furthermore, a comprehensive segmentation approach is developed, tailored to the microdamage features of silicon nitride bearings. In this process, the sigma function is optimized to eliminate the influence of zero-value elements, effectively reducing image noise and improving overall image quality. Additionally, iterative refinement of contour lines is implemented to ensure accurate delineation of actual microdamage boundaries, thereby enabling precise segmentation of microdamage images. The images processed with dynamic scale median filtering and the active contour iteration model exhibit a peak signal-to-noise ratio of 44.66 dB, an average structural similarity of up to 97.63 %, an average extraction accuracy of 98.07 %, and an average edge overlap of 93.68 %, effectively reducing the impact of artifact contours and edge collapse features on the extraction of damaged images. This method has improved the efficiency of failure analysis for silicon nitride bearings, provided a safeguard for the development of image processing technology, ensured the safe use of silicon nitride bearings in the field of new energy vehicles, and indirectly promoted the development of the new energy sector.