diff --git a/AlgorithmModule/CMakeLists.txt b/AlgorithmModule/CMakeLists.txt
index f6299c7..f25d151 100644
--- a/AlgorithmModule/CMakeLists.txt
+++ b/AlgorithmModule/CMakeLists.txt
@@ -46,6 +46,7 @@ ${PROJECT_SOURCE_DIR}/ConfigModule/include
 ${PROJECT_SOURCE_DIR}/Common/include
 ${PROJECT_SOURCE_DIR}/AIEngineModule/include
 ${PROJECT_SOURCE_DIR}/AIEngineModule/include_base
+${PROJECT_SOURCE_DIR}/TcsCheckModule/include
 )
 link_directories(
 /usr/local/lib/
@@ -67,6 +68,7 @@ add_library(TY_Check SHARED ${SRC_LISTS})
 target_link_libraries(TY_Check
 	nvinfer
 	Config
+	TcsCheck
 	${OpenCV_LIBS}
 	${CUDA_LIBRARIES}
 	)
diff --git a/AlgorithmModule/include/ImgCheckAnalysisy.hpp b/AlgorithmModule/include/ImgCheckAnalysisy.hpp
index a4f754d..b9acbe5 100644
--- a/AlgorithmModule/include/ImgCheckAnalysisy.hpp
+++ b/AlgorithmModule/include/ImgCheckAnalysisy.hpp
@@ -36,6 +36,7 @@
 #include "AI_Factory.h"
 #include "ImageAllResult.h"
 #include "Task.h"
+#include "TcsCheck.h"
 
 using namespace std;
 using namespace cv;
@@ -263,6 +264,9 @@ private:
     Edge_QX_Det::DetConfigResult m_Edge_DetConfig;
     std::vector<QXImageResult> m_Draw_qxImageResult; // 缺陷小图结果
 
+    // 传统检测模块
+    CTcsCheck m_tcsCheck;
+
     std::vector<cv::Rect> SmallRoiList;
 
     std::string m_strRootPath_TA_cls;
diff --git a/AlgorithmModule/src/ImgCheckAnalysisy.cpp b/AlgorithmModule/src/ImgCheckAnalysisy.cpp
index 0506c30..3450871 100644
--- a/AlgorithmModule/src/ImgCheckAnalysisy.cpp
+++ b/AlgorithmModule/src/ImgCheckAnalysisy.cpp
@@ -1611,14 +1611,33 @@ int ImgCheckAnalysisy::Traditional_Detect_Thread(const cv::Mat &img, cv::Mat &Re
 	std::string strBaseLog = "Traditional_Detect";
 	m_pdetlog->AddCheckstr(PrintLevel_0, DET_LOG_LEVEL_3, strBaseLog, "Traditional_Detect  Start");
 
-	// 输出与AI推理相同格式的二值mask图 (CV_8UC1)
-	ResultImg = cv::Mat::zeros(img.size(), CV_8UC1);
-
-	// ===== TODO: 在此处调用传统检测算法库 =====
-	// 示例：cv::threshold(img, ResultImg, 128, 255, cv::THRESH_BINARY);
-	// 输入:  img (单通道灰度图, CV_8UC1)
-	// 输出:  ResultImg (二值mask, CV_8UC1, 0/255)
-	// ==========================================
+	// 首次调用时初始化传统检测参数（从 m_AnalysisyConfig 映射）
+	static bool bTcsInited = false;
+	if (!bTcsInited)
+	{
+		CHECK_PARAM cp;
+		cp.nAreaLowFilter  = 80;
+		cp.nBlockSize      = 100;
+		cp.nDiscardTop     = 170;
+		cp.nDiscardBottom  = 170;
+		cp.nDiscardLeft    = 180;
+		cp.nDiscardRight   = 460;
+		cp.fZoomRatio      = 0.25f;
+		cp.nFilterLow      = 15;
+		cp.nFilterHigh     = 15;
+		cp.nAreaFilter     = 10;
+		cp.nCountFilter    = 50;
+		m_tcsCheck.SetChecConfig(&cp);
+		bTcsInited = true;
+	}
+
+	// 调用传统检测：输出残点二值图
+	int ret = m_tcsCheck.TraditionalDetect(img, ResultImg);
+	if (ret != 0)
+	{
+		m_pdetlog->AddCheckstr(PrintLevel_0, DET_LOG_LEVEL_3, strBaseLog, "Traditional_Detect  FAILED (no product)");
+		return -1;
+	}
 
 	m_pdetlog->AddCheckstr(PrintLevel_0, DET_LOG_LEVEL_3, strBaseLog, "Traditional_Detect  End");
 	return 0;
@@ -1792,8 +1811,32 @@ int ImgCheckAnalysisy::AI_QX_Class_Thread()
 // ======================== 传统分类 ========================
 int ImgCheckAnalysisy::Traditional_QX_Class_Thread()
 {
-	m_pdetlog->AddCheckstr(PrintLevel_1, DET_LOG_LEVEL_3, "Traditional_Class", " Start (placeholder)");
+	m_pdetlog->AddCheckstr(PrintLevel_1, DET_LOG_LEVEL_3, "Traditional_Class", " Start");
+
+	// 1. 调用传统分类：基于上次 TraditionalDetect 缓存的模糊图 + 当前 mask
+	if (m_pImageAllResult == nullptr || m_pImageAllResult->AIMaskImg.empty())
+	{
+		m_pdetlog->AddCheckstr(PrintLevel_1, DET_LOG_LEVEL_3, "Traditional_Class", " No mask image");
+		return -1;
+	}
+
+	int nDefectCount = m_tcsCheck.TraditionalClassify(m_pImageAllResult->AIMaskImg);
+	if (nDefectCount < 0)
+	{
+		m_pdetlog->AddCheckstr(PrintLevel_1, DET_LOG_LEVEL_3, "Traditional_Class", " Classify failed");
+		return -1;
+	}
+
+	// 2. TcsCheck 缺陷类型 → CONFIG_QX_NAME 映射表
+	static const int TcsDefectToConfigQX[] = {
+		CONFIG_QX_NAME_cell_other,       // DEFECT_TYPE_OK = 0
+		CONFIG_QX_NAME_cell_dianzhuang,  // DEFECT_TYPE_POINT = 1    (硬质颗粒 → 点状)
+		CONFIG_QX_NAME_cell_line,        // DEFECT_TYPE_SCRATCH = 2  (划伤 → 线状)
+		CONFIG_QX_NAME_cell_zangwu,      // DEFECT_TYPE_DIRTY = 3    (脏污)
+		CONFIG_QX_NAME_cell_danban,      // DEFECT_TYPE_FADING_SPOTS = 4 (淡斑)
+	};
 
+	// 3. 将分类结果匹配到 blobs.blobTab（基于位置/面积最近邻匹配）
 	int totalTasks = blobs.blobCount;
 	for (int i = 0; i < totalTasks; i++)
 	{
@@ -1802,19 +1845,45 @@ int ImgCheckAnalysisy::Traditional_QX_Class_Thread()
 		if (pblob->ErrType == ERR_TYPE_2)
 		{
 			pblob->AIclasstype = CONFIG_QX_NAME_cell_ymhs;
+			continue;
+		}
+
+		// 在 TcsCheck 结果中找最佳匹配（中心距离最近）
+		int bestIdx = -1;
+		int bestDist2 = INT_MAX;
+		int blobCenterX = (pblob->minx + pblob->maxx) / 2;
+		int blobCenterY = (pblob->miny + pblob->maxy) / 2;
+
+		for (int j = 0; j < nDefectCount; j++)
+		{
+			const DEFECT_INFO& info = m_tcsCheck.m_vecDefectInfo[j];
+			int defCenterX = info.nDefectX + info.nDefectWidth / 2;
+			int defCenterY = info.nDefectY + info.nDefectHeight / 2;
+			int dx = blobCenterX - defCenterX;
+			int dy = blobCenterY - defCenterY;
+			int dist2 = dx * dx + dy * dy;
+
+			// 面积接近的优先（容差 50% 内）
+			int areaDiff = std::abs(pblob->area - info.nDefectArea);
+			if (areaDiff < info.nDefectArea / 2 && dist2 < bestDist2)
+			{
+				bestDist2 = dist2;
+				bestIdx = j;
+			}
+		}
+
+		if (bestIdx >= 0)
+		{
+			int defectType = m_tcsCheck.m_vecDefectInfo[bestIdx].nDefectType;
+			pblob->AIclasstype = TcsDefectToConfigQX[defectType];
 		}
 		else
 		{
-			// ===== TODO: 在此处调用传统分类算法库 =====
-			// 输入:  pblob->minx/maxy/miny/maxy 定位的blob区域
-			//        m_pImageAllResult->detImg 原始检测图
-			// 输出:  pblob->AIclasstype (CONFIG_QX_NAME_cell_xxx)
-			// ==========================================
 			pblob->AIclasstype = CONFIG_QX_NAME_cell_other;
 		}
 	}
 
-	m_pdetlog->AddCheckstr(PrintLevel_1, DET_LOG_LEVEL_3, "Traditional_Class", " End (placeholder), classified %d blobs", totalTasks);
+	m_pdetlog->AddCheckstr(PrintLevel_1, DET_LOG_LEVEL_3, "Traditional_Class", " End, classified %d/%d blobs", nDefectCount, totalTasks);
 	return 0;
 }
 
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 1808a0e..18d6197 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -19,7 +19,7 @@ message(STATUS "oPENCV Library status:")
 message(STATUS ">version:${OpenCV_VERSION}")
 message(STATUS "Include:${OpenCV_INCLUDE_DIRS}")
 
-set(CMAKE_BUILD_TYPE "debug")
+set(CMAKE_BUILD_TYPE "release")
 set(CMAKE_CXX_STANDARD 17)
 set(CMAKE_CXX_STANDARD_REQUIRED ON)  # 要求编译器支持 C++17
 set(CMAKE_CXX_EXTENSIONS OFF)        # 禁用 GNU 特定的扩展
diff --git a/TcsCheckModule/include/TcsCheck.h b/TcsCheckModule/include/TcsCheck.h
index a214a2b..df33bb4 100644
--- a/TcsCheckModule/include/TcsCheck.h
+++ b/TcsCheckModule/include/TcsCheck.h
@@ -95,6 +95,15 @@ public:
     void SetChecConfig(CHECK_PARAM* cp);
     void ProcessImages(bool bDrawResult);
 
+    // ========== 独立检测/分类接口 ==========
+    // 传统检测：对单张图做完整预处理+自适应二值化，输出残点二值图
+    // 返回: 0=成功, -1=无产品/输入为空
+    int TraditionalDetect(const cv::Mat& img, cv::Mat& blobImg);
+    // 传统分类：对残点二值图做连通域分析+缺陷分类，结果写入 m_vecDefectInfo
+    // 前提: 已调用 TraditionalDetect（内部 m_matBlur 已就绪）
+    // 返回: 分类到的缺陷数量，<0 表示异常
+    int TraditionalClassify(const cv::Mat& blobImg);
+
 private:
     
     std::string m_strDirIn;
@@ -102,6 +111,7 @@ private:
 
     cv::Mat m_matLoad;
     cv::Mat m_matBlob;
+    cv::Mat m_matBlur;     // 模糊图，供 TraditionalClassify 使用
     cv::Mat m_matDraw;
 
     cv::Size m_sizeImage;
diff --git a/TcsCheckModule/src/TcsCheck.cpp b/TcsCheckModule/src/TcsCheck.cpp
index b4aea37..dcaf55b 100644
--- a/TcsCheckModule/src/TcsCheck.cpp
+++ b/TcsCheckModule/src/TcsCheck.cpp
@@ -309,6 +309,67 @@ cv::Mat CTcsCheck::DrawBlobInfoImage(const cv::Mat& imgCrop, const cv::Mat& imgB
 
 	return cropColor;
 }
+// ============================================================
+// TraditionalDetect — 独立检测接口
+// 对单张图做：阈值定位→裁剪→缩放→高斯模糊→自适应二值化
+// 输出残点二值图 blobImg，内部缓存 m_matBlur 供后续分类
+// 返回: 0=成功, -1=无产品或输入异常
+// ============================================================
+int CTcsCheck::TraditionalDetect(const cv::Mat& img, cv::Mat& blobImg)
+{
+	if (img.empty()) return -1;
+
+	m_matLoad = img;
+	m_sizeImage = img.size();
+
+	// 1. 全局阈值 → 产品区域定位
+	cv::Mat matBinary;
+	cv::threshold(m_matLoad, matBinary, m_cpCfg.nAreaLowFilter, 255, cv::THRESH_BINARY);
+
+	// 2. 获取最大连通域外接矩形
+	cv::Rect rtValid = GetBoundingRect(matBinary);
+	if (rtValid == cv::Rect(0, 0, 0, 0))
+	{
+		blobImg = cv::Mat();
+		return -1;
+	}
+
+	// 3. 裁剪边缘
+	cv::Rect rtCrop = GetCropArea(rtValid);
+	cv::Mat matCrop = m_matLoad(rtCrop).clone();
+
+	// 4. 缩放
+	const int outW = std::max(1, static_cast<int>(matCrop.cols * m_cpCfg.fZoomRatio));
+	const int outH = std::max(1, static_cast<int>(matCrop.rows * m_cpCfg.fZoomRatio));
+	cv::Mat matResized;
+	cv::resize(matCrop, matResized, cv::Size(outW, outH), 0, 0, cv::INTER_AREA);
+
+	// 5. 高斯模糊 — 缓存到 m_matBlur
+	cv::GaussianBlur(matResized, m_matBlur, cv::Size(5, 5), 0);
+
+	// 6. 自适应二值化检测
+	m_matBlob = AdaptiveBinary(m_matBlur);
+	blobImg = m_matBlob.clone();
+
+	return 0;
+}
+
+// ============================================================
+// TraditionalClassify — 独立分类接口
+// 对残点二值图做连通域分析+缺陷分类决策树
+// 前提: 已调用 TraditionalDetect（m_matBlur 已缓存）
+// 返回: 分类到的缺陷数量，<0 表示异常
+// ============================================================
+int CTcsCheck::TraditionalClassify(const cv::Mat& blobImg)
+{
+	if (m_matBlur.empty() || blobImg.empty()) return -1;
+
+	m_matBlob = blobImg.clone();
+	ClassifyBlobs(m_matBlur, blobImg);
+
+	return static_cast<int>(m_vecDefectInfo.size());
+}
+
 void CTcsCheck::ProcessImages(bool bDrawResult)
 {
     LoadImages();