机器视觉检测设备图像预处理的方式有哪些？

由于获取条件的不同和外界的各种干扰，经过CCD视觉成像系统采集到的原始图像往往存在着大量的噪声和失真，这种数据无法直接用于机器视觉检测系统。为了消除外界环境对图像采集的干扰，需要对图像进行预处理，例如通过图像分析和识别等手段，消除使图像质量恶化的因素，使采集到的图像能够更有效的用于有效信息的提取，接下来，我们就讲一讲机器视觉检测设备图像预处理的三种方式。

Due to the different acquisition conditions various external disturbances, the original images collected by CCD vision imaging system often have a lot of noise distortion, this data can be directly used in the machine vision detection system. In order to eliminate the interference of the external environment to the image acquisition, it is necessary to preprocess the image, for example, by means of image analysis recognition, to eliminate the factors that worsen the image quality, so that the captured image can be used more effectively for the extraction of effective information, then we will talk about three ways of image preprocessing of machine vision detection equipment.

均值滤波:其是一种线性滤波算法，用图片中目标像素周围8个像素的平均值来代替该像素自身，从而达到降噪效果。但是该算法自身存在一定的缺陷，会破环图像的细节部分，使其变得模糊，不能有效的去除噪点。

Mean filtering: It is a linear filtering algorithm, which replaces the pixel itself with the average value of 8 pixels around the target pixel in the picture, so as to achieve the noise reduction effect. However, the algorithm itself has some defects, will break the detail part of the image, so that it becomes blurred, can effectively remove the noise point.

中值滤波:是一种基于统计排序理论的非线性滤波算法，其将待处理的像素点用周围的8个或者24个像素点中的中值进行替换，从而达到降噪的目的。

Median filtering: is a nonlinear filtering algorithm based on statistical sorting theory, which replaces the pixel points to be processed with the median values in the surrounding 8 24 pixels, so as to achieve the goal of noise reduction.

高斯滤波:其为一种线性平滑滤波算法，用于处理高斯噪声，将待处理的像素点用周围其他像素点的加权平均值代替。高斯滤波处理对于服从正态分布的噪声特别有效。

Gaussian filtering: It is a linear smoothing filtering algorithm for processing Gaussian noise, replacing the pixel points to be processed with the weighted average values of other pixels around them. Gaussian filter processing is particularly effective for noise subject to normal distribution.