图像处理中的滑动窗口概念 [英] Sliding window concept in Image processing

查看:1928
本文介绍了图像处理中的滑动窗口概念的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

海,
我正在寻找在图像上实现滑动窗口的概念.
如何选择窗口以及更好的尺寸.我正在处理256x256尺寸的图像.
或者至少有人可以解释如何选择滑动窗口并在矩阵(2-d数组)上实现

Hai,
I am looking for implementing sliding window concept on images.
How to select the window and what size is preferable.I am working on 256x256 size image.
Or atleast someone could explain how to choose sliding window and implement on a matrix(2-d array)

推荐答案

我假设您正在谈论像image这样的内核窗口操作过滤器.窗口的大小取决于目的,内核函数需要多少个最小样本以及图像的对象分辨率.

例如,假设您的图像上有5 x 5像素区域的小圆圈.如果您的窗口大小为5 x 5,则在使用均值函数进行平滑操作时,这些圆将被平滑.但是,如果您的窗口大小为10 x 10,则圆圈可能会消失.因此,是否要删除圆圈取决于目的.

因此,窗口大小取决于纯粹的特定目的,一般规则可能并不总是有效.
I assume you are talking about the kernel window operations like image filters. The size of the window depends on the purpose, how many minimum samples your kernel function needs and the object resolution of the image.

For example assume your image has small small circles of 5 x 5 pixel area. If your windows size is 5 x 5, in a smoothing operation using a mean function then those circles get smoothed. But if your window size is 10 x 10 then the circles may disappear. So it depends on the purpose whether you want remove the circles or not.

So the window size depends on purely specific purpose, a general rule may not work always.


这篇关于图像处理中的滑动窗口概念的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆