基于图像融合技术的大豆图像滤波方法研究

摘要

Based on machine vision of appearance quality detection of soybean is a hotspot research in recent years, the filtering of soybean image is one of the important work of appearance quality detection of soybean .In order to remove the noise of soybean image better, this research puts forward a kind of soybean image filtering method based on image fusion. The algorithm put Wiener filter and morphological filter on a sample image respectively, then go on image fusion algo-rithm based on wavelet transform, effectively solve the image edge burr phenomenon, the experiment proves that this method guarantees the integrity of details and the edge of the image , and good effect of filtering .%基于机器视觉的大豆外观品质检测一直是近年来研究的热点，其中大豆图像的滤波是大豆外观品质检测的重要工作内容之一。为了更好地去除大豆图像的噪声，提出了一种基于图像融合技术的大豆图像滤波方法。该算法对一个样本图像分别进行维纳滤波和形态学滤波，在此基础上进行基于小波变换的图像融合算法，有效解决了图像边缘毛刺现象。实验证明，此方法保证了图像的细节和边缘的完整性，图像滤波效果良好。