diff --git "a/docs/quickStart/4-4\346\250\241\345\236\213Tensor\346\225\260\346\215\256\345\244\204\347\220\206&\350\207\252\345\256\232\344\271\211\346\250\241\345\236\213\345\220\216\345\244\204\347\220\206.md" "b/docs/quickStart/4-4\346\250\241\345\236\213Tensor\346\225\260\346\215\256\345\244\204\347\220\206&\350\207\252\345\256\232\344\271\211\346\250\241\345\236\213\345\220\216\345\244\204\347\220\206.md"
deleted file mode 100644
index 2d705c404e8145463905eb5b6dd7a5b5ccdad919..0000000000000000000000000000000000000000
--- "a/docs/quickStart/4-4\346\250\241\345\236\213Tensor\346\225\260\346\215\256\345\244\204\347\220\206&\350\207\252\345\256\232\344\271\211\346\250\241\345\236\213\345\220\216\345\244\204\347\220\206.md"
+++ /dev/null
@@ -1,122 +0,0 @@
-# 模型Tensor数据处理&插件后处理
-
-本章节将指导模型Tensor数据解析和封装相关操作。
-配套样例描述如何自定义一个插件进行后处理操作。
-****
-**注意！**  
-本样例中后处理指使用模型输出的原始metadata，自行开发插件来进行后处理。  
-当类型为4-2中相关的内置类型时，效率不如后处理so库方式
-****
-
-## medadata结构说明
-该结构为模型推理插件mxpi_tensorinfer输出的原始数据，按以下层级封装。
-
-1. MxpiTensorPackageList
-模型tensor组合列表。
-```protobuf
-repeated MxpiTensorPackage tensorPackageVec;
-```
-2. MxpiTensorPackage
-
-模型tensor组合数据结。
-```protobuf
-repeated MxpiMetaHeader headerVec;
-repeated MxpiTensor tensorVec;
-```
-3. MxpiTensor
-
-模型tensor数据结构。
-```protobuf
-uint64 tensorDataPtr; // 内存指针数值
-int32 tensorDataSize; // 内存大小，需要和实际内存大小一致，否则可能会导致coredump
-uint32 deviceId; // Device编号
-MxpiMemoryType memType; // 内存类型
-uint64 freeFunc; // 内存销毁函数指针
-repeated int32 tensorShape; // 张量形状
-bytes dataStr; // 内存中的数据
-int32 tensorDataType; //内存中张量的数据类型
-```
-
-## C++结构体说明
-该结构体为TensorBase数据结构相关说明，详情可参考TensorBase.h头文件。  
-（位于%SDK%/include/MxBase/Tensor/TensorBase/）  
-本处仅摘抄部分关键信息
-```c++
-enum TensorDataType {
-    TENSOR_DTYPE_UNDEFINED = -1,
-    TENSOR_DTYPE_FLOAT32 = 0,
-    TENSOR_DTYPE_FLOAT16 = 1,
-    TENSOR_DTYPE_INT8 = 2,
-    TENSOR_DTYPE_INT32 = 3,
-    TENSOR_DTYPE_UINT8 = 4,
-    TENSOR_DTYPE_INT16 = 6,
-    TENSOR_DTYPE_UINT16 = 7,
-    TENSOR_DTYPE_UINT32 = 8,
-    TENSOR_DTYPE_INT64 = 9,
-    TENSOR_DTYPE_UINT64 = 10,
-    TENSOR_DTYPE_DOUBLE64 = 11,
-    TENSOR_DTYPE_BOOL = 12
-};
-    // 获取tensor部署的设备类型
-    MemoryData::MemoryType GetTensorType() const;
-    // buffer记录的数据量
-    size_t GetSize() const;
-    // buffer 字节数据量
-    size_t GetByteSize() const;
-    // tensor 的shape
-    std::vector<uint32_t> GetShape() const;
-    std::vector<uint32_t> GetStrides() const;
-    // tensor 的 device
-    int32_t GetDeviceId() const;
-    // tensor 数据类型
-    TensorDataType GetDataType() const;
-    uint32_t GetDataTypeSize() const;
-    // 判断是否在Host
-    bool IsHost() const;
-    // 判断是否在Device
-    bool IsDevice() const;
-    // 获取tensor指针
-    void* GetBuffer() const;
-    APP_ERROR GetBuffer(void *&ptr, const std::vector<uint32_t> &indices) const;
-    // host to device
-    APP_ERROR ToDevice(int32_t deviceId);
-    // host to dvpp
-    APP_ERROR ToDvpp(int32_t deviceId);
-    // device to host
-    APP_ERROR ToHost();
-    // 组batch相关
-    static APP_ERROR BatchConcat(const std::vector<TensorBase> &inputs, TensorBase &output);
-    static APP_ERROR BatchStack(const std::vector<TensorBase> &inputs, TensorBase &output);
-    static APP_ERROR BatchVector(const std::vector<TensorBase> &inputs, TensorBase &output,
-        const bool &keepDims = false);
-    // 详细信息
-    std::string GetDesc();
-    // 检查错误
-    APP_ERROR CheckTensorValid() const;
-```
-## 样例说明
-参考[4-1插件开发调试指导](4-1插件开发调试指导.md)部署自定义插件样例，示例使用samplePluginPostProc作为工程名，远程目录名同样为samplePluginPostProc。  
-- 更改mxVision/C++/main.cpp中94行所使用的pipeline为样例中的SamplePluginPost.pipeline
-- 更改mxVision/python/main.py中32行使用的pipeline为样例中的SamplePluginPost.pipeline
-- 相比4-1样例中的SamplePlugin.pipeline，本样例中pipeline使用新后处理框架下的模型推理插件mxpi_tensorinfer输出原始Tensor至自定义插件并完成后处理示例。
->该文件位于样例根目录，但代码中实际指向mxVision/pipeline文件夹下，这是为了与使用原有Sample.pipeline的样例统一目录。实际使用时复制pipeline文件或更改代码中的路径均可  
-
-****
-本样例部署测试同4-1，不再重复。
-****
-**运行结果**
-1. C++样例  
-
->使用菜单项Build/Rebuild All ....后运行该程序，如果无错误可在输出中看到类似的以下字段  其中MxpiSamplePlugin.cpp文件中的日志输出为插件内部结果，main.cpp日志输出为stream结果。此处tensor[0]相关值应该相同
-```shell
-I0730 10:36:55.459653 11981 MxpiSamplePlugin.cpp:118] MxpiSamplePlugin::Process start
-W0730 10:36:55.459954 11981 MxpiSamplePlugin.cpp:98] source Tensor number:3
-W0730 10:36:55.460223 11981 MxpiSamplePlugin.cpp:99] Tensor[0] ByteSize in .cpp:172380
-I0730 10:36:55.462554 11981 MxpiSamplePlugin.cpp:174] MxpiSamplePlugin::Process end
-I0730 10:36:55.462568 11935 main.cpp:129] Results:{"MxpiClass":[{"classId":42,"className":"The shape of tensor[0] in metadata is 172380, Don’t panic!","confidence":0.314}]}
-```
-2. Python样例  
->修改python样例目录下的run.sh脚本（该脚本用于配置环境变量），确认MX_SDK_HOME环境变量指向正确的位置，自定义插件路径已位于GST_PLUGIN_PATH中，或手动修改环境变量，然后执行程序。正常情况下输出以下类似字段
-```shell
-{"MxpiClass":[{"classId":42,"className":"The shape of tensor[0] in metadata is 172380, Don’t panic!","confidence":0.314}]}
-```
\ No newline at end of file
diff --git a/tutorials/SamplePostProcess/src/samplepostprocess/README.md b/tutorials/SamplePostProcess/src/samplepostprocess/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..bdbe2398fc2a38a8200713ae17dde1440d881df1
--- /dev/null
+++ b/tutorials/SamplePostProcess/src/samplepostprocess/README.md
@@ -0,0 +1,8 @@
+# Resnset50样例
+本项目(tutorials\SamplePostProcess\src\samplepostprocess)构建了resnet50网络的新版后处理插件。
+## 继承类型
+>MxBase/PostProcessBases/ClassPostProcessBase.h
+## 使用说明
+使用mxpi_classpostprocessor加载后处理so
+
+>参考https://support.huawei.com/enterprise/zh/doc/EDOC1100207069/4e71a146
\ No newline at end of file
diff --git a/tutorials/SamplePostProcess/src/vggpostprocess/CMakeLists.txt b/tutorials/SamplePostProcess/src/vggpostprocess/CMakeLists.txt
new file mode 100644
index 0000000000000000000000000000000000000000..b6c01eb54a60c62f0a1234d13640d6093fac5333
--- /dev/null
+++ b/tutorials/SamplePostProcess/src/vggpostprocess/CMakeLists.txt
@@ -0,0 +1,27 @@
+cmake_minimum_required(VERSION 3.5.2)
+project(vggpostprocess)
+
+add_definitions(-D_GLIBCXX_USE_CXX11_ABI=0 -Dgoogle=mindxsdk_private)
+
+set(PLUGIN_NAME "vggpostprocess")
+set(TARGET_LIBRARY ${PLUGIN_NAME})
+set(LIBRARY_OUTPUT_PATH ${PROJECT_SOURCE_DIR}/../../lib)
+
+include_directories(${MX_SDK_HOME}/include)
+include_directories(${MX_SDK_HOME}/opensource/include)
+include_directories(${MX_SDK_HOME}/opensource/include/gstreamer-1.0)
+include_directories(${MX_SDK_HOME}/opensource/include/glib-2.0)
+include_directories(${MX_SDK_HOME}/opensource/lib/glib-2.0/include)
+
+link_directories(${MX_SDK_HOME}/opensource/lib)
+link_directories(${MX_SDK_HOME}/lib)
+
+add_compile_options(-std=c++11 -fPIC -fstack-protector-all -pie -Wno-deprecated-declarations)
+add_compile_options("-DPLUGIN_NAME=${PLUGIN_NAME}")
+
+add_definitions(-DENABLE_DVPP_INTERFACE)
+add_library(${TARGET_LIBRARY} SHARED vggPostProcess.cpp vggPostProcess.h)
+
+target_link_libraries(${TARGET_LIBRARY} glib-2.0 gstreamer-1.0 gobject-2.0 gstbase-1.0 gmodule-2.0 glog)
+target_link_libraries(${TARGET_LIBRARY} plugintoolkit mxpidatatype mxbase)
+#target_link_libraries(${TARGET_LIBRARY} -Wl,-z,relro,-z,now,-z,noexecstack -s)
diff --git a/tutorials/SamplePostProcess/src/vggpostprocess/README.md b/tutorials/SamplePostProcess/src/vggpostprocess/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..dd9f9c575b8e1b225e20a14225241e98372e125f
--- /dev/null
+++ b/tutorials/SamplePostProcess/src/vggpostprocess/README.md
@@ -0,0 +1,8 @@
+# ssdvgg样例
+本项目(tutorials\SamplePostProcess\src\vggpostprocess)构建了vggpostprocess网络的新版后处理插件。
+## 继承类型
+>MxBase/PostProcessBases/ObjectPostProcessBase.h
+## 使用说明
+需补完后处理中tensor检查和object处理逻辑部分，使用mxpi_objectpostprocessor加载后处理so
+
+>参考https://support.huawei.com/enterprise/zh/doc/EDOC1100207069/397691d0
\ No newline at end of file
diff --git a/tutorials/SamplePostProcess/src/vggpostprocess/vggPostProcess.cpp b/tutorials/SamplePostProcess/src/vggpostprocess/vggPostProcess.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..987cc487405681a5ab8a62c9ed4ce5d6bf79198c
--- /dev/null
+++ b/tutorials/SamplePostProcess/src/vggpostprocess/vggPostProcess.cpp
@@ -0,0 +1,115 @@
+/*
+ * Copyright(C) 2020. Huawei Technologies Co.,Ltd. All rights reserved.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include "VggPostProcess.h"
+#include "MxBase/Log/Log.h"
+
+namespace {
+/// need manunel set
+}
+namespace MxBase {
+    VggPostProcess &VggPostProcess::operator=(const VggPostProcess &other) {
+        if (this == &other) {
+            return *this;
+        }
+        ObjectPostProcessBase::operator=(other);
+        return *this;
+    }
+
+    APP_ERROR VggPostProcess::Init(const std::map <std::string, std::shared_ptr<void>> &postConfig) {
+        LogDebug << "Start to Init VggPostProcess.";
+        APP_ERROR ret = ObjectPostProcessBase::Init(postConfig);
+        if (ret != APP_ERR_OK) {
+            LogError << GetError(ret) << "Fail to superInit in ObjectPostProcessBase.";
+            return ret;
+        }
+        LogDebug << "End to Init VggPostProcess.";
+        return APP_ERR_OK;
+    }
+
+    APP_ERROR VggPostProcess::DeInit() {
+        return APP_ERR_OK;
+    }
+
+    bool VggPostProcess::IsValidTensors(const std::vector <TensorBase> &tensors) const {
+    /// need write check
+    }
+
+    void VggPostProcess::ObjectDetectionOutput(const std::vector <TensorBase> &tensors,
+                                                  std::vector <std::vector<ObjectInfo>> &objectInfos,
+                                                  const std::vector <ResizedImageInfo> &resizedImageInfos) {
+        LogDebug << "VggPostProcess start to write results.";
+        for (auto num : {objectNumTensor_,objectInfoTensor_}) {
+            if ((num >= tensors.size()) || (num <0)) {
+                LogError << GetError(APP_ERR_INVALID_PARAM) << "TENSOR(" << num 
+                         << ") must ben less than tensors'size(" << tensors.size() << ") and larger than 0.";
+            }
+        }
+        uint32_t batchSize = tensors[objectNumTensor_].GetShape()[0];
+        for (uint32_t i = 0; i < batchSize; i++) {
+            std::vector <ObjectInfo> objectInfo;
+            int* objectNumPtr = (int*)GetBuffer(tensors[objectInfoTensor_],i);
+            float* objectInfoPtr = (float*)GetBuffer(tensors[objectInfoTensor_],i);
+            for (int j = 0; j< *objectNumPtr; j++) {
+                ObjectInfo objInfo;
+                /// need write objInfo
+            }
+            }
+            objectInfos.push_back(objectInfo);
+        }
+        LogDebug << "VggPostProcess write results successed.";
+    }
+
+    APP_ERROR VggPostProcess::Process(const std::vector <TensorBase> &tensors,
+                                         std::vector <std::vector<ObjectInfo>> &objectInfos,
+                                         const std::vector <ResizedImageInfo> &resizedImageInfos,
+                                         const std::map <std::string, std::shared_ptr<void>> &paramMap) {
+        LogDebug << "Start to Process VggPostProcess.";
+        APP_ERROR ret = APP_ERR_OK;
+        if (resizedImageInfos.size() == 0) {
+            ret = APP_ERR_INPUT_NOT_MATCH;
+            LogError << GetError(ret) << "resizedImageInfos is not provided which is necessary for VggPostProcess.";
+            return ret;
+        }
+        auto inputs = tensors;
+        ret = CheckAndMoveTensors(inputs);
+        if (ret != APP_ERR_OK) {
+            LogError << GetError(ret) << "CheckAndMoveTensors failed.";
+            return ret;
+        }
+
+        ObjectDetectionOutput(inputs, objectInfos, resizedImageInfos);
+
+        for (uint32_t i = 0; i < resizedImageInfos.size(); i++) {
+            CoordinatesReduction(i, resizedImageInfos[i], objectInfos[i]);
+        }
+        LogObjectInfos(objectInfos);
+        LogDebug << "End to Process VggPostProcess.";
+        return APP_ERR_OK;
+    }
+
+    extern "C" {
+    std::shared_ptr <MxBase::VggPostProcess> GetObjectInstance() {
+        LogInfo << "Begin to get VggPostProcess instance.";
+        auto instance = std::make_shared<MxBase::VggPostProcess>();
+        if (instance - nullptr) {
+            LogError << "Creat VggPostProcess object failed. Failed to allocate memory.";
+        }
+        LogInfo << "End to get VggPostProcess instance.";
+        return instance;
+    }
+    }
+}
\ No newline at end of file
diff --git a/tutorials/SamplePostProcess/src/vggpostprocess/vggPostProcess.h b/tutorials/SamplePostProcess/src/vggpostprocess/vggPostProcess.h
new file mode 100644
index 0000000000000000000000000000000000000000..f60f6f9e6bed52eb62ae1620efd9724955086048
--- /dev/null
+++ b/tutorials/SamplePostProcess/src/vggpostprocess/vggPostProcess.h
@@ -0,0 +1,62 @@
+/*
+ * Copyright(C) 2020. Huawei Technologies Co.,Ltd. All rights reserved.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef VGG_POST_PROCESS_H
+#define VGG_POST_PROCESS_H
+#include "MxBase/PostProcessBases/ObjectPostProcessBase.h"
+
+namespace {
+const int DEFAULT_OBJECT_NUM_TENSOR = 0;
+const int DEFAULT_OBJECT_INFO_TENSOR = 1;
+}
+
+namespace MxBase {
+    class VggPostProcess{
+
+    public:
+        VggPostProcess() = default;
+
+        ~VggPostProcess() = default;
+
+        VggPostProcess(const VggPostProcess &other) = default;
+
+        VggPostProcess &operator=(const VggPostProcess &other);
+
+        APP_ERROR Init(const std::map <std::string, std::shared_ptr<void>> &postConfig) override;
+
+        APP_ERROR DeInit() override;
+
+        APP_ERROR Process(const std::vector <MxBase::TensorBase> &tensors,
+                          std::vector <std::vector<MxBase::ObjectInfo>> &objectInfos,
+                          const std::vector <MxBase::ResizedImageInfo> &resizedImageInfos = {},
+                          const std::map <std::string, std::shared_ptr<void>> &paramMap = {}) override;
+
+    protected:
+        bool IsValidTensors(const std::vector <MxBase::TensorBase> &tensors) const override;
+
+        void ObjectDetectionOutput(const std::vector <MxBase::TensorBase> &tensors,
+                                   std::vector <std::vector<MxBase::ObjectInfo>> &objectInfos,
+                                   const std::vector <MxBase::ResizedImageInfo> &resizedImageInfos = {});
+    private:
+        uint32_t objectNumTensor_ = DEFAULT_OBJECT_NUM_TENSOR;
+        uint32_t objectInfoTensor_ = DEFAULT_OBJECT_INFO_TENSOR;
+
+    };
+    extern "C" {
+    std::shared_ptr<MxBase::VggPostProcess> GetObjectInstance();
+    }
+}
+#endif
\ No newline at end of file
diff --git a/tutorials/SamplePostProcess/src/yolov3/README.md b/tutorials/SamplePostProcess/src/yolov3/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..5f0af5c0c45a0d9df635735b8b23a3b877360e0e
--- /dev/null
+++ b/tutorials/SamplePostProcess/src/yolov3/README.md
@@ -0,0 +1,9 @@
+# yolov3样例
+
+本项目(tutorials\SamplePostProcess\src\yolov3)构建了yolov3网络的新版后处理插件。
+## 继承类型
+>MxBase/PostProcessBases/ObjectPostProcessBase.h
+## 使用说明
+使用mxpi_objectpostprocessor加载后处理so
+
+>参考https://support.huawei.com/enterprise/zh/doc/EDOC1100207069/397691d0
\ No newline at end of file
diff --git a/tutorials/SamplePostProcess/src/yolov3/Yolov3PostProcess.h b/tutorials/SamplePostProcess/src/yolov3/Yolov3PostProcess.h
index 9e7db513caf64b1a01650b16fb692ddb114b1318..30f847794e46a59548757e73f78ccf16a5f38f82 100644
--- a/tutorials/SamplePostProcess/src/yolov3/Yolov3PostProcess.h
+++ b/tutorials/SamplePostProcess/src/yolov3/Yolov3PostProcess.h
@@ -45,7 +45,7 @@ struct NetInfo {
 }
 
 namespace MxBase {
-    class Yolov3PostProcess{
+    class Yolov3PostProcess: public ObjectPostProcessBase{
 
     public:
         Yolov3PostProcess() = default;