摘要
针对采集到的浮选泡沫图像存在光照缺失、噪声干扰以及边缘不明显等问题,提出一种基于对比度受限的局部直方图均衡(CLAHE)和三维块匹配滤波(BM3D)结合的浮选泡沫图像预处理方法。首先,将输入的泡沫图像切分为多个图块,对图块的直方图进行均衡计算并限制其对比度。然后,通过BM3D算法利用图像自身的信息对增强之后的图像进行降噪处理。结果表明:处理后的图像质量得到了明显的提升,4种典型工况的浮选泡沫图像的峰值信噪比(PSNR)相比几种经典算法分别提升了33.64%、15.40%、11.83%和11.7%,结构相似性(SSIM)分别提升了11.89%、7.28%、4.09%和5%。
Aiming at the problems of lack of illumination, noise interference and indistinct edges in flotation froth images collected, a pretreatment method for flotation froth images based on contrast limited local histogram equalization(CLAHE) and 3D block matching filtering(BM3D) is proposed.First, the input foam image is segmented into several blocks, and the histogram of the block is calculated and the contrast is restricted. Then, the enhanced image is denoised by using the information of the image itself through BM3D algorithm. The results show that the image quality has been improved significantly. The peak signal to noise ratio(PSNR) of flotation froth images in four typical operating conditions has increased by 33.64%, 15.40%, 11.83% and 11.7% compared with several classical algorithms,and the structure similarity(SSIM) has increased 11.89%, 7.28%, 4.09% and 5% respectively.
作者
王宇龙
王然风
付翔
曹文艳
WANG Yulong;WANG Ranfeng;FU Xiang;CAO Wenyan(College of Mining Engineering,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《煤炭技术》
CAS
北大核心
2022年第6期222-224,共3页
Coal Technology
基金
山西省应用基础研究计划重点自然基金(201901D111007)
山西省关键核心技术和共性技术研发攻关专项项目(2020XXX004)。
关键词
浮选泡沫
图像增强
图像去噪
机器视觉
flotation foam
image enhancement
image denoising
machine vision