摘要
在焊接条件下，用工业摄像机获取的焊缝的实时图像中常含有各种干扰信息，导致传统的图像边缘检测的方法很难准确地检测出焊缝的边缘。为了提高焊缝实时图像处理的准确性，提出了宏观处理与微观检测相结合的思想。分形理论被引入了图像处理，建立了焊缝图像的分形数学模型，对其性质进行了讨论与证明。最后给出了图像综合处理的方法。这种图像处理思想不仅可用于焊接图像处理，对图像处理向智能处理方向发展也有重要的启发意义。
Under welding condition, it is much noisy in weld seam image that is obtained by industrial camera in real time. It is difficult to detect the edges of weld seam correctly by using traditional image edge detecting method. To process the weld seam image exactly, an idea has been proposed of combination of macroscopical processing with microcosmic detecting. Fractal theory is introdused into image processing, and the weld seam image fractal model is then established. The properties of the model are discussed and proved. At last, a comprehensive method of image processing is released. This method is not only used in welding image processing, but also significant for the development of intellectual image processing.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
1999年第2期23-26,共4页
Journal of Tsinghua University(Science and Technology)
基金
国家自然科学基金
关键词
焊缝图像
实时处理
分形模型
灰度强度
焊接
weld seam image
real time processing
fractal model
gray intensity
