基于改进CLAHE算法的复杂场景下熔池轮廓分割
Molten Pool Contour Segmentation in Complex Scenes Based on Improved CLAHE Algorithm
朱心愿 1曾俊冬 1宋小军 1毛月1
作者信息
- 1. 上海电力大学 电子与信息工程学院,上海 200135
- 折叠
摘要
针对包含钨极、送丝杆、金属氧化物等多要素的复杂场景下熔池图像对比度较低、干扰区域过多,致使熔池轮廓难以分割的问题,本文将CLAHE(限制对比度自适应直方图均衡化)算法应用到熔池图像处理领域.首先对图像去噪,然后采用一种基于改进CLAHE算法的熔池图像分割算法对图像进行增强处理,然后利用Otsu算法对增强后的图像进行分割.该改进CLAHE算法方法结合了CLAHE算法和伽马校正,克服了原始图像对比度低,干扰区域多的缺陷.随机选取了 3 幅复杂场景下的熔池图像进行了处理,并与其他算法进行对比.结果表明,本文所提算法在增强效果和分割效果上均有明显提升,不仅改善了熔池边缘对比度,还抑制了干扰区域,提高了熔池轮廓分割的准确性和完整性,有助于自动焊精确识别熔池边缘,防止出现咬边,确保焊接质量.
Abstract
Aiming at the problem that it is difficult to segment the molten pool contour due to low contrast and too many interference areas of the molten pool image in complex scenes containing tungsten pole,wire feed rod,metal oxide and other elements,the CLAHE(limited contrast adaptive histogram equalization)algorithm was applied to the molten pool image processing field.Firstly,the image was denoised,then an image segmentation method of molten pool based on improved CLAHE algorithm was applied to enhance the image.Finally,the Otsu algorithm was used to segment the enhanced image.The improved CLAHE algorithm combines CLAHE algorithm and gamma correction to overcome the defects of low contrast and many interference areas of the original image.Three images of molten pool in complex scenes were randomly selected for processing objects,and it was compared with other algorithms.The results show that the proposed algorithm has significantly improved enhancement effect and the segmentation results.It not only improves the molten pool edge contrast,but also suppresses the interference area,improves the accuracy and completeness of the molten pool contour segmentation,which is helpful to accurately identify the molten pool edge during automatic welding,prevent undercut and ensure the welding quality.
关键词
复杂场景/图像增强/CLAHE算法/熔池轮廓分割Key words
complex scene/image enhancement/CLAHE algorithm/molten pool contour segmentation引用本文复制引用
基金项目
国家自然科学基金项目(62105196)
中国博士后科学基金面上项目(2019M661334)
出版年
2024
NETL
NSTL
万方数据