文章作者：激萌の小宅

促销：￥0

价格：￥0

配送方式：购买后立即生效（如购买异常，请联系站长）

付款之后一定要等待自动跳转结束，否则购买可能会失败

0 天

有效期
0

总销量
0

累计评价

详细说明

<h1 style="font-size: 32px; border-bottom: 2px solid rgb(204, 204, 204); padding: 0px 4px 0px 0px; margin: 0px 0px 10px;">BRIEF描述符</h1>当前系列所有demo下载地址： <a href="https://github.com/GaoRenBao/OpenCv4-Demo" target="_blank" style="font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; margin: 0px; padding: 0px; list-style: none; border: 0px; color: rgb(0, 102, 0); transition-duration: 0.2s; transition-property: opacity; outline: none; opacity: 0.8;">https://github.com/GaoRenBao/OpenCv4-Demo</a><p style="margin-top: 0px; margin-bottom: 10px; box-sizing: border-box; overflow-wrap: break-word; word-break: break-all; line-height: 1.5em; font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; color: rgb(51, 51, 51); text-wrap: wrap; ">源码参考原文：<p style="margin-top: 0px; margin-bottom: 10px; box-sizing: border-box; overflow-wrap: break-word; word-break: break-all; line-height: 1.5em; font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; color: rgb(51, 51, 51); text-wrap: wrap; "><a href="/api/system/download?file=OpenCV-Python-Tutorial-%E4%B8%AD%E6%96%87%E7%89%88.pdf" target="_blank" title="OpenCV-Python-Tutorial-中文版.pdf （P205）">OpenCV-Python-Tutorial-中文版.pdf （P205）</a>  <p style="margin-top: 0px; margin-bottom: 10px; text-wrap: wrap; padding: 0px; list-style: none; border: 0px; overflow-wrap: break-word; word-break: break-all; line-height: 1.5em; font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; box-sizing: border-box; color: rgb(51, 51, 51);">不同编程语言对应的OpenCv版本以及开发环境信息如下： <table border="1" style="margin-left: 2px; border-right: none; border-bottom: none; border-image: initial; border-left: 1px solid rgb(102, 102, 102); border-top: 1px solid rgb(102, 102, 102);"><tbody><tr class="firstRow"><td width="81" valign="top" style="border-width: 1px; border-color: windowtext rgb(102, 102, 102) rgb(102, 102, 102) windowtext; border-bottom-style: solid; border-right-style: solid; padding: 5px;">语言</td><td width="424" valign="top" style="border-width: 1px; border-color: windowtext rgb(102, 102, 102) rgb(102, 102, 102) windowtext; border-bottom-style: solid; border-right-style: solid; padding: 5px;">OpenCv版本</td><td width="321" valign="top" style="border-width: 1px; border-color: windowtext rgb(102, 102, 102) rgb(102, 102, 102) windowtext; border-bottom-style: solid; border-right-style: solid; padding: 5px;">IDE</td></tr><tr><td width="81" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">C#</td><td width="424" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">OpenCvSharp4.4.8.0.20230708</td><td width="321" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">Visual Studio 2022</td></tr><tr><td width="81" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">C++</td><td width="424" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">OpenCv-4.5.5-vc14_vc15这个版本缺少xfeatures2d库）OpenCv-4.6.0（这个版本缺少xfeatures2d库）</td><td width="321" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">Visual Studio 2022</td></tr><tr><td width="81" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">Python</td><td width="424" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">OpenCv-Python (4.6.0.66)OpenCv-Python (4.5.5)（这个版本缺少xfeatures2d库）</td><td width="321" valign="top" style="border-left-width: 1px; border-left-color: windowtext; border-top: none; border-bottom: 1px solid rgb(102, 102, 102); border-right: 1px solid rgb(102, 102, 102); padding: 5px;">PyCharm Community Edition 2022.1.3</td></tr></tbody></table><p style="margin-top: 0px; margin-bottom: 10px; box-sizing: border-box; overflow-wrap: break-word; word-break: break-all; line-height: 1.5em; font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; color: rgb(51, 51, 51); text-wrap: wrap; ">         算法使用的是已经平滑后的图像        非常重要的一点是:BRIEF是-种特征描述符,它不提供查找特征的方法。所以我们不得不使用其他特征检测器，比如SIFT和SURF等。原始文献推荐使用CenSurE特征检测器，这种算法很快。而且BRIEF算法对CenSurE关键点的描述效果要比SURF关键点的描述更好。        简单来说BRIEF是一种对特征点描述符计算和匹配的快速方法。这种算法可以实现很高的识别率，除非出现平面内的大旋转。        下面的代码使用了CenSurE特征检测器和BRIEF描述符。(在OpenCV中CenSurE检测器被叫做STAR检测器)。 测试原图如下： <img src="/upload/image/6383453188145459191230459.jpg" title="blox.jpg" alt="blox.jpg"/> <strong style="margin: 0px; padding: 0px; list-style: none; border: 0px; color: rgb(51, 51, 51); font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; ">C#版本代码如下：<strong style="margin: 0px; padding: 0px; list-style: none; border: 0px; color: rgb(51, 51, 51); font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; "><strong style="margin: 0px; padding: 0px; list-style: none; border: 0px; color: rgb(51, 51, 51); font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; "><img src="/upload/image/6383453653421862878536175.jpg" title="QQ截图20231102154213.jpg" alt="QQ截图20231102154213.jpg"/><pre class="brush:c#;toolbar:false">using OpenCvSharp; using System; namespace ConsoleApp {     internal class Program     {         static void Main(string[] args)         {             Mat img = Cv2.ImRead("../../../images/blox.jpg", 0);             // 创建BRIEF描述符提取器             var brief = OpenCvSharp.XFeatures2D.BriefDescriptorExtractor.Create();             // Initiate FAST detector             var star = OpenCvSharp.XFeatures2D.StarDetector.Create();             var kp = star.Detect(img);             // compute the descriptors with BRIEF             Mat des = new Mat();             brief.Compute(img, ref kp, des);             Console.WriteLine(brief.DescriptorSize);             Console.WriteLine($"{des.Width},{des.Height}");         }     } }</pre> <strong style="text-wrap: wrap; margin: 0px; padding: 0px; list-style: none; border: 0px; color: rgb(51, 51, 51); font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; ">C++版本代码如下：<pre class="brush:cpp;toolbar:false">【缺XFeatures2D算法】</pre> <strong style="text-wrap: wrap; margin: 0px; padding: 0px; list-style: none; border: 0px; color: rgb(51, 51, 51); font-family: "Helvetica Neue", Helvetica, "PingFang SC", "Hiragino Sans GB", "Microsoft YaHei", "Noto Sans CJK SC", "WenQuanYi Micro Hei", Arial, sans-serif; ">Python版本代码如下：<img src="/upload/image/6383453733905760408504277.jpg" title="QQ截图20231102155535.jpg" alt="QQ截图20231102155535.jpg"/><pre class="brush:python;toolbar:false">import cv2 img = cv2.imread('../images/blox.jpg', 0) # 创建BRIEF描述符提取器 brief = cv2.xfeatures2d.BriefDescriptorExtractor_create() # Initiate FAST detector star = cv2.xfeatures2d.StarDetector_create() # Initiate BRIEF extractor brief = cv2.xfeatures2d.BriefDescriptorExtractor_create() # find the keypoints with STAR kp = star.detect(img, None) # compute the descriptors with BRIEF kp, des = brief.compute(img, kp) print(brief.descriptorSize()) print(des.shape)</pre>

BRIEF描述符

当前系列所有demo下载地址：

https://github.com/GaoRenBao/OpenCv4-Demo

源码参考原文：

OpenCV-Python-Tutorial-中文版.pdf （P205）

不同编程语言对应的OpenCv版本以及开发环境信息如下：

语言	OpenCv版本	IDE
C#	OpenCvSharp4.4.8.0.20230708	Visual Studio 2022
C++	OpenCv-4.5.5-vc14_vc15这个版本缺少xfeatures2d库） OpenCv-4.6.0（这个版本缺少xfeatures2d库）	Visual Studio 2022
Python	OpenCv-Python (4.6.0.66) OpenCv-Python (4.5.5)（这个版本缺少xfeatures2d库）	PyCharm Community Edition 2022.1.3

算法使用的是已经平滑后的图像

非常重要的一点是:BRIEF是-种特征描述符,它不提供查找特征的方法。所以我们不得不使用其他特征检测器，比如SIFT和SURF等。原始文献推荐使用CenSurE特征检测器，这种算法很快。而且BRIEF算法对CenSurE关键点的描述效果要比SURF关键点的描述更好。

简单来说BRIEF是一种对特征点描述符计算和匹配的快速方法。这种算法可以实现很高的识别率，除非出现平面内的大旋转。

下面的代码使用了CenSurE特征检测器和BRIEF描述符。(在OpenCV中CenSurE检测器被叫做STAR检测器)。

测试原图如下：

C#版本代码如下：

using OpenCvSharp;
using System;

namespace ConsoleApp
{
    internal class Program
    {
        static void Main(string[] args)
        {
            Mat img = Cv2.ImRead("../../../images/blox.jpg", 0);

            // 创建BRIEF描述符提取器
            var brief = OpenCvSharp.XFeatures2D.BriefDescriptorExtractor.Create();

            // Initiate FAST detector
            var star = OpenCvSharp.XFeatures2D.StarDetector.Create();
            var kp = star.Detect(img);

            // compute the descriptors with BRIEF
            Mat des = new Mat();
            brief.Compute(img, ref kp, des);
            Console.WriteLine(brief.DescriptorSize);
            Console.WriteLine($"{des.Width},{des.Height}");
        }
    }
}

C++版本代码如下：

【缺XFeatures2D算法】

Python版本代码如下：

QQ截图20231102155535.jpg

import cv2

img = cv2.imread('../images/blox.jpg', 0)

# 创建BRIEF描述符提取器
brief = cv2.xfeatures2d.BriefDescriptorExtractor_create()

# Initiate FAST detector
star = cv2.xfeatures2d.StarDetector_create()

# Initiate BRIEF extractor
brief = cv2.xfeatures2d.BriefDescriptorExtractor_create()

# find the keypoints with STAR
kp = star.detect(img, None)
# compute the descriptors with BRIEF
kp, des = brief.compute(img, kp)

print(brief.descriptorSize())
print(des.shape)