From 6cc26faeab2de4bd68f52455af035fce77a779dd Mon Sep 17 00:00:00 2001
From: zjc-zjc-123 <1714105370@qq.com>
Date: Mon, 11 Nov 2024 17:57:08 +0800
Subject: [PATCH] =?UTF-8?q?=E8=BF=81=E7=A7=BB=E9=A1=B9=E7=9B=AE=E5=88=B0CL?=
 =?UTF-8?q?ion?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .gitignore           |   4 +-
 CMakeLists.txt       |  26 +++++++
 README.md            |  13 +++-
 cotton_color.cpp     | 146 +++++++++++++++++++++++-----------------
 cotton_color.vcxproj |   7 +-
 cotton_color2.cpp    | 157 +++++++++++++++++++++++++++++++++++++++++++
 6 files changed, 286 insertions(+), 67 deletions(-)
 create mode 100644 CMakeLists.txt
 create mode 100644 cotton_color2.cpp

diff --git a/.gitignore b/.gitignore
index 9491a2f..621be39 100644
--- a/.gitignore
+++ b/.gitignore
@@ -360,4 +360,6 @@ MigrationBackup/
 .ionide/
 
 # Fody - auto-generated XML schema
-FodyWeavers.xsd
\ No newline at end of file
+FodyWeavers.xsd
+
+.idea
\ No newline at end of file
diff --git a/CMakeLists.txt b/CMakeLists.txt
new file mode 100644
index 0000000..b60ffde
--- /dev/null
+++ b/CMakeLists.txt
@@ -0,0 +1,26 @@
+cmake_minimum_required(VERSION 3.29)
+project(cotton_color)
+
+
+set(CMAKE_CXX_STANDARD 17)
+
+# 设置 OpenCV 路径
+set(OpenCV_DIR "E:/opencv4.10/opencv/build")
+
+# 找到 OpenCV 包
+find_package(OpenCV REQUIRED)
+
+add_definitions(-DUNICODE -D_UNICODE)
+# 包含头文件路径
+include_directories(${OpenCV_INCLUDE_DIRS})
+
+# 添加可执行文件
+add_executable(cotton_color cotton_color.cpp)
+# 链接 OpenCV 库
+target_link_libraries(cotton_color ${OpenCV_LIBS}  comdlg32)
+
+
+# 添加可执行文件
+add_executable(cotton_color2 cotton_color2.cpp)
+# 链接 OpenCV 库
+target_link_libraries(cotton_color2 ${OpenCV_LIBS}  comdlg32)
\ No newline at end of file
diff --git a/README.md b/README.md
index 29c2b25..8053519 100644
--- a/README.md
+++ b/README.md
@@ -49,5 +49,14 @@ L a* b* 色彩空间检测，检测明黄色、白色。
 									激进方案    			 -> 深度学习 - > 区块判别 -> 杂质     |
 																					      |
 									保守方案											   -> 深度学习确认 -> 杂质
-															
-															 
\ No newline at end of file
+
+讨论记录：
+
+暗红色（棉叶）
+
+棉花亮度低
+
+黑线、黑色孔洞
+
+土黄
+
diff --git a/cotton_color.cpp b/cotton_color.cpp
index 877fc36..f8624bb 100644
--- a/cotton_color.cpp
+++ b/cotton_color.cpp
@@ -1,135 +1,157 @@
-#include <opencv2/opencv.hpp>
+#include <opencv2/opencv.hpp>
 #include <iostream>
 #include <map>
 #include <string>
 #include <windows.h>
-#include <commdlg.h>  // 包含文件对话框相关的函数
-
+#include <commdlg.h>  // �����ļ��Ի�����صĺ���
 
 using namespace cv;
 using namespace std;
 
-
 /**
- * @brief 鲜艳色彩检测函数，通过饱和度阈值检测输入图像中鲜艳的颜色区域。
+ * @brief ����ɫ�ʼ�⺯����ͨ�����Ͷ���ֵ�������ͼ�������޵���ɫ����
  *
- * 此函数将输入图像从 BGR 色彩空间转换到 HSV 色彩空间，并提取出饱和度 (S) 通道。然后，通过设置饱和度阈值，
- * 来检测图像中饱和度大于阈值的区域，标记为输出图像的鲜艳颜色区域。
- *
- * @param inputImage 输入图像，类型为 cv::Mat，要求为 BGR 色彩空间。
- * @param outputImage 输出图像，类型为 cv::Mat，输出图像将标记出鲜艳颜色区域，原始图像尺寸。
- * @param saturationThreshold 饱和度阈值，类型为 int，用于过滤低饱和度的区域，范围通常为 0 到 255。
- *                            饱和度高于此阈值的区域将被认为是鲜艳的颜色。
- *
- * @note 饱和度阈值越高，输出图像将只保留更为鲜艳的区域。
- *       若饱和度阈值过低，可能会检测到过多的区域。
+ * @param inputImage ����ͼ������Ϊ cv::Mat��Ҫ��Ϊ BGR ɫ�ʿռ䡣
+ * @param outputImage ���ͼ������Ϊ cv::Mat�����ͼ�񽫱�dz�������ɫ����ԭʼͼ��ߴ硣
+ * @param params ����ӳ�䣬���ڴ��ݸ��ֿ����õIJ������籥�Ͷ���ֵ�ȡ�
  */
- /**
-  * @brief 鲜艳色彩检测函数，通过饱和度阈值检测输入图像中鲜艳的颜色区域。
-  *
-  * @param inputImage 输入图像，类型为 cv::Mat，要求为 BGR 色彩空间。
-  * @param outputImage 输出图像，类型为 cv::Mat，输出图像将标记出鲜艳颜色区域，原始图像尺寸。
-  * @param params 参数映射，用于传递各种可配置的参数，如饱和度阈值等。
-  */
 void vibrantColorDetection(const Mat& inputImage, Mat& outputImage, const map<string, int>& params) {
-    // 从参数映射中获取饱和度阈值
+    // �Ӳ���ӳ���л�ȡ���Ͷ���ֵ
     int saturationThreshold = params.at("saturationThreshold");
 
-    // 将输入图像从 BGR 转换为 HSV
+    // ������ͼ��� BGR ת��Ϊ HSV
     Mat hsvImage;
     cvtColor(inputImage, hsvImage, COLOR_BGR2HSV);
 
-    // 分离 HSV 图像的各个通道
+    // ���� HSV ͼ��ĸ���ͨ��
     Mat channels[3];
     split(hsvImage, channels);
 
-    // 获取饱和度通道 (S)
+    // ��ȡ���Ͷ�ͨ�� (S)
     Mat saturation = channels[1];
 
-    // 创建输出图像，将饱和度大于阈值的区域标记为杂质
+    // �������ͼ�񣬽����Ͷȴ�����ֵ��������Ϊ����
     outputImage = Mat::zeros(inputImage.size(), CV_8UC1);
 
-    // 对饱和度图像应用阈值处理
+    // �Ա��Ͷ�ͼ��Ӧ����ֵ����
     threshold(saturation, outputImage, saturationThreshold, 255, THRESH_BINARY);
 }
 
+/**
+ * @brief ���ļ��Ի��򣬷���ѡ���ļ���·����
+ *
+ * @return ѡ���ļ�������·��������Ϊ std::wstring������û�ȡ��ѡ�񣬷��ؿ��ַ�����
+ */
+std::wstring openFileDialog() {
+    // ��ʼ���ļ�ѡ��Ի���
+    OPENFILENAMEW ofn;       // ʹ�ÿ��ַ��汾�Ľṹ
+    wchar_t szFile[260] = {0};    // �洢ѡ����ļ�·��
 
-string openFileDialog() {
-    // 初始化文件选择对话框
-    OPENFILENAME ofn;       // 文件对话框结构
-    wchar_t szFile[260];    // 存储选择的文件路径
-
-    // 设置 OPENFILENAME 结构的默认值
+    // ���� OPENFILENAMEW �ṹ��Ĭ��ֵ
     ZeroMemory(&ofn, sizeof(ofn));
     ofn.lStructSize = sizeof(ofn);
     ofn.hwndOwner = NULL;
-    ofn.lpstrFile = szFile;
+    ofn.lpstrFile = szFile;  // �����ļ�·��������
     ofn.nMaxFile = sizeof(szFile) / sizeof(szFile[0]);
     ofn.lpstrFilter = L"Image Files\0*.BMP;*.JPG;*.JPEG;*.PNG;*.GIF\0All Files\0*.*\0";
     ofn.nFilterIndex = 1;
-    ofn.lpstrFileTitle = NULL;
+    ofn.lpstrFileTitle = NULL;  // ����Ҫ�������ļ���
     ofn.nMaxFileTitle = 0;
-    ofn.lpstrInitialDir = NULL;
-    ofn.lpstrTitle = L"Select an image file";
+    ofn.lpstrInitialDir = NULL;  // ʹ��Ĭ�ϳ�ʼĿ¼
+    ofn.lpstrTitle = L"Select an image file";  // �Ի������
     ofn.Flags = OFN_PATHMUSTEXIST | OFN_FILEMUSTEXIST;
 
-    // 打开文件选择对话框
-    if (GetOpenFileName(&ofn) == TRUE) {
-        // 将 wchar_t 转换为 string
-        wstring ws(szFile);
-        string filePath(ws.begin(), ws.end());
-        return filePath;
+    // ���ļ�ѡ��Ի���
+    if (GetOpenFileNameW(&ofn) == TRUE) {
+        return szFile;  // ����ѡ�е��ļ�·��
     }
 
-    return "";  // 如果用户取消，返回空字符串
+    return L"";  // ����û�ȡ�������ؿ��ַ���
 }
 
-
+/**
+ * @brief ��ȡͼ���ļ���֧�� Unicode ·����
+ *
+ * @return ���ص�ͼ������Ϊ cv::Mat���������ʧ�ܣ����ؿյ� Mat��
+ */
 Mat readImage() {
-    // 读取输入图像
-    string imagePath = openFileDialog();
+    // ��ȡ����ͼ��·��
+    std::wstring imagePath = openFileDialog();
 
     if (imagePath.empty()) {
-        cout << "No file selected or user cancelled." << endl;
+        wcout << L"No file selected or user cancelled." << endl;
         return Mat();
     }
 
-    // 使用 OpenCV 读取选中的图片
-    Mat image = imread(imagePath);
+    // ʹ�� Windows API ���ļ�
+    HANDLE hFile = CreateFileW(imagePath.c_str(), GENERIC_READ, FILE_SHARE_READ, NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);
+    if (hFile == INVALID_HANDLE_VALUE) {
+        wcout << L"Error: Could not open file." << endl;
+        return Mat();
+    }
+
+    // ��ȡ�ļ���С
+    LARGE_INTEGER fileSize;
+    if (!GetFileSizeEx(hFile, &fileSize)) {
+        wcout << L"Error: Could not get file size." << endl;
+        CloseHandle(hFile);
+        return Mat();
+    }
+
+    if (fileSize.QuadPart > MAXDWORD) {
+        wcout << L"Error: File size too large." << endl;
+        CloseHandle(hFile);
+        return Mat();
+    }
+
+    DWORD dwFileSize = static_cast<DWORD>(fileSize.QuadPart);
+
+    // ��ȡ�ļ����ݵ�������
+    std::vector<BYTE> buffer(dwFileSize);
+    DWORD bytesRead = 0;
+    if (!ReadFile(hFile, buffer.data(), dwFileSize, &bytesRead, NULL) || bytesRead != dwFileSize) {
+        wcout << L"Error: Could not read file." << endl;
+        CloseHandle(hFile);
+        return Mat();
+    }
+
+    CloseHandle(hFile);
+
+    // ʹ�� OpenCV ���ڴ滺������ȡͼ��
+    Mat image = imdecode(buffer, IMREAD_COLOR);
 
     if (image.empty()) {
-        cout << "Error: Could not load image." << endl;
+        wcout << L"Error: Could not decode image." << endl;
         return Mat();
     }
 
     return image;
 }
 
-
 int main() {
-    // 读取输入图像
-	Mat inputImage = readImage();
+    // ��ȡ����ͼ��
+    Mat inputImage = readImage();
 
     if (inputImage.empty()) {
         cout << "Error: Could not load image." << endl;
         return -1;
     }
 
-    // 创建输出图像
+    // �������ͼ��
     Mat outputImage;
 
-    // 使用 map 模拟 JSON 参数传递
+    // ʹ�� map ģ���������
     map<string, int> params;
-    params["saturationThreshold"] = 100;  // 设置饱和度阈值为100
+    params["saturationThreshold"] = 100;  // ���ñ��Ͷ���ֵΪ 100
 
-    // 调用鲜艳颜色检测函数
+    // ����������ɫ��⺯��
     vibrantColorDetection(inputImage, outputImage, params);
 
-    // 显示原图和检测到的鲜艳区域
+    // ��ʾԭͼ�ͼ�⵽����������
     imshow("Original Image", inputImage);
     imshow("Detected Vibrant Colors", outputImage);
 
-    // 等待用户按键
+    // �ȴ��û�����
     waitKey(0);
     return 0;
-}
\ No newline at end of file
+}
diff --git a/cotton_color.vcxproj b/cotton_color.vcxproj
index fd94705..713fece 100644
--- a/cotton_color.vcxproj
+++ b/cotton_color.vcxproj
@@ -104,10 +104,13 @@
       <SDLCheck>true</SDLCheck>
       <PreprocessorDefinitions>_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
       <ConformanceMode>true</ConformanceMode>
+      <AdditionalIncludeDirectories>$(mil_path64)\..\include;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
       <GenerateDebugInformation>true</GenerateDebugInformation>
+      <AdditionalLibraryDirectories>$(mil_path64)\..\lib\</AdditionalLibraryDirectories>
+      <AdditionalDependencies>mil.lib;Mil3d.lib;milim.lib;miledge.lib;%(AdditionalDependencies)</AdditionalDependencies>
     </Link>
   </ItemDefinitionGroup>
   <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
@@ -118,14 +121,14 @@
       <SDLCheck>true</SDLCheck>
       <PreprocessorDefinitions>NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
       <ConformanceMode>true</ConformanceMode>
-      <AdditionalIncludeDirectories>C:\Users\ZLSDKJ\source\opencv\build\include;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
+      <AdditionalIncludeDirectories>C:\Users\zjc\source\opencv\build\include;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
     </ClCompile>
     <Link>
       <SubSystem>Console</SubSystem>
       <EnableCOMDATFolding>true</EnableCOMDATFolding>
       <OptimizeReferences>true</OptimizeReferences>
       <GenerateDebugInformation>true</GenerateDebugInformation>
-      <AdditionalLibraryDirectories>C:\Users\ZLSDKJ\source\opencv\build\x64\vc16\lib;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
+      <AdditionalLibraryDirectories>C:\Users\zjc\source\opencv\build\x64\vc16\lib;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
       <AdditionalDependencies>Comdlg32.lib;opencv_world4100.lib;%(AdditionalDependencies)</AdditionalDependencies>
     </Link>
   </ItemDefinitionGroup>
diff --git a/cotton_color2.cpp b/cotton_color2.cpp
new file mode 100644
index 0000000..c6ab671
--- /dev/null
+++ b/cotton_color2.cpp
@@ -0,0 +1,157 @@
+#include <opencv2/opencv.hpp>
+#include <iostream>
+#include <map>
+#include <string>
+#include <windows.h>
+#include <commdlg.h>  // 包含文件对话框相关的函数
+
+using namespace cv;
+using namespace std;
+
+/**
+ * @brief 鲜艳色彩检测函数，通过饱和度阈值检测输入图像中鲜艳的颜色区域。
+ *
+ * @param inputImage 输入图像，类型为 cv::Mat，要求为 BGR 色彩空间。
+ * @param outputImage 输出图像，类型为 cv::Mat，输出图像将标记出鲜艳颜色区域，原始图像尺寸。
+ * @param params 参数映射，用于传递各种可配置的参数，如饱和度阈值等。
+ */
+void vibrantColorDetection(const Mat& inputImage, Mat& outputImage, const map<string, int>& params) {
+    // 从参数映射中获取饱和度阈值
+    int saturationThreshold = params.at("saturationThreshold");
+
+    // 将输入图像从 BGR 转换为 HSV
+    Mat hsvImage;
+    cvtColor(inputImage, hsvImage, COLOR_BGR2HSV);
+
+    // 分离 HSV 图像的各个通道
+    Mat channels[3];
+    split(hsvImage, channels);
+
+    // 获取饱和度通道 (S)
+    Mat saturation = channels[1];
+
+    // 创建输出图像，将饱和度大于阈值的区域标记为杂质
+    outputImage = Mat::zeros(inputImage.size(), CV_8UC1);
+
+    // 对饱和度图像应用阈值处理
+    threshold(saturation, outputImage, saturationThreshold, 255, THRESH_BINARY);
+}
+
+/**
+ * @brief 打开文件对话框，返回选中文件的路径。
+ *
+ * @return 选中文件的完整路径，类型为 std::wstring。如果用户取消选择，返回空字符串。
+ */
+std::wstring openFileDialog() {
+    // 初始化文件选择对话框
+    OPENFILENAMEW ofn;       // 使用宽字符版本的结构
+    wchar_t szFile[260] = {0};    // 存储选择的文件路径
+
+    // 设置 OPENFILENAMEW 结构的默认值
+    ZeroMemory(&ofn, sizeof(ofn));
+    ofn.lStructSize = sizeof(ofn);
+    ofn.hwndOwner = NULL;
+    ofn.lpstrFile = szFile;  // 设置文件路径缓冲区
+    ofn.nMaxFile = sizeof(szFile) / sizeof(szFile[0]);
+    ofn.lpstrFilter = L"Image Files\0*.BMP;*.JPG;*.JPEG;*.PNG;*.GIF\0All Files\0*.*\0";
+    ofn.nFilterIndex = 1;
+    ofn.lpstrFileTitle = NULL;  // 不需要单独的文件名
+    ofn.nMaxFileTitle = 0;
+    ofn.lpstrInitialDir = NULL;  // 使用默认初始目录
+    ofn.lpstrTitle = L"Select an image file";  // 对话框标题
+    ofn.Flags = OFN_PATHMUSTEXIST | OFN_FILEMUSTEXIST;
+
+    // 打开文件选择对话框
+    if (GetOpenFileNameW(&ofn) == TRUE) {
+        return szFile;  // 返回选中的文件路径
+    }
+
+    return L"";  // 如果用户取消，返回空字符串
+}
+
+/**
+ * @brief 读取图像文件，支持 Unicode 路径。
+ *
+ * @return 加载的图像，类型为 cv::Mat。如果加载失败，返回空的 Mat。
+ */
+Mat readImage() {
+    // 读取输入图像路径
+    std::wstring imagePath = openFileDialog();
+
+    if (imagePath.empty()) {
+        wcout << L"No file selected or user cancelled." << endl;
+        return Mat();
+    }
+
+    // 使用 Windows API 打开文件
+    HANDLE hFile = CreateFileW(imagePath.c_str(), GENERIC_READ, FILE_SHARE_READ, NULL, OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);
+    if (hFile == INVALID_HANDLE_VALUE) {
+        wcout << L"Error: Could not open file." << endl;
+        return Mat();
+    }
+
+    // 获取文件大小
+    LARGE_INTEGER fileSize;
+    if (!GetFileSizeEx(hFile, &fileSize)) {
+        wcout << L"Error: Could not get file size." << endl;
+        CloseHandle(hFile);
+        return Mat();
+    }
+
+    if (fileSize.QuadPart > MAXDWORD) {
+        wcout << L"Error: File size too large." << endl;
+        CloseHandle(hFile);
+        return Mat();
+    }
+
+    DWORD dwFileSize = static_cast<DWORD>(fileSize.QuadPart);
+
+    // 读取文件内容到缓冲区
+    std::vector<BYTE> buffer(dwFileSize);
+    DWORD bytesRead = 0;
+    if (!ReadFile(hFile, buffer.data(), dwFileSize, &bytesRead, NULL) || bytesRead != dwFileSize) {
+        wcout << L"Error: Could not read file." << endl;
+        CloseHandle(hFile);
+        return Mat();
+    }
+
+    CloseHandle(hFile);
+
+    // 使用 OpenCV 从内存缓冲区读取图像
+    Mat image = imdecode(buffer, IMREAD_COLOR);
+
+    if (image.empty()) {
+        wcout << L"Error: Could not decode image." << endl;
+        return Mat();
+    }
+
+    return image;
+}
+
+int main() {
+    // 读取输入图像
+    Mat inputImage = readImage();
+
+    if (inputImage.empty()) {
+        cout << "Error: Could not load image." << endl;
+        return -1;
+    }
+
+    // 创建输出图像
+    Mat outputImage;
+
+    // 使用 map 模拟参数传递
+    map<string, int> params;
+    params["saturationThreshold"] = 100;  // 设置饱和度阈值为 100
+
+    // 调用鲜艳颜色检测函数
+    vibrantColorDetection(inputImage, outputImage, params);
+
+    // 显示原图和检测到的鲜艳区域
+    imshow("Original Image", inputImage);
+    imshow("Detected Vibrant Colors", outputImage);
+
+    // 等待用户按键
+    waitKey(0);
+    return 0;
+}