[Model] update pptracking letterboxresize and add some comment

examples/vision/tracking/pptracking/README.md

@@ -0,0 +1,35 @@
+# PP-Tracking模型部署
+
+## 模型版本说明
+
+- [PaddleDetection release/2.5](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5)
+
+## 支持模型列表
+
+目前FastDeploy支持如下模型的部署
+
+- [PP-Tracking系列模型](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/configs/mot)
+
+
+## 导出部署模型
+
+在部署前，需要先将训练好的PP-Tracking导出成部署模型，导出PPTracking导出模型步骤，参考文档[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/deploy/pptracking/cpp/README.md)。
+
+
+## 下载预训练模型
+
+为了方便开发者的测试，下面提供了PP-Tracking行人跟踪垂类模型，开发者可直接下载使用，更多模型参见[PPTracking](https://github.com/PaddlePaddle/PaddleDetection/blob/release/2.5/deploy/pptracking/README_cn.md)。
+
+| 模型                                                                                                   | 参数大小   | 精度    | 备注 |
+|:-----------------------------------------------------------------------------------------------------|:-------|:----- | :------ |
+| [PP-Tracking](https://bj.bcebos.com/paddlehub/fastdeploy/fairmot_hrnetv2_w18_dlafpn_30e_576x320.tgz) | 51.2MB | - |
+
+**说明**
+- 仅支持JDE模型（JDE，FairMOT，MCFairMOT)；
+- 目前暂不支持SDE模型的部署，待PaddleDetection官方更新SED部署代码后，对SDE模型进行支持。
+
+
+## 详细部署文档
+
+- [Python部署](python)
+- [C++部署](cpp)

examples/vision/tracking/pptracking/cpp/README.md

@@ -22,7 +22,6 @@ make -j
 wget https://bj.bcebos.com/paddlehub/fastdeploy/fairmot_hrnetv2_w18_dlafpn_30e_576x320.tgz
 tar -xvf fairmot_hrnetv2_w18_dlafpn_30e_576x320.tgz
 wget https://bj.bcebos.com/paddlehub/fastdeploy/person.mp4
-wget https://bj.bcebos.com/paddlehub/fastdeploy/person.mp4
 
 
 # CPU推理

fastdeploy/vision/common/processors/letter_box.cc

@@ -0,0 +1,71 @@
+// Copyright (c) 2022 PaddlePaddle Authors. 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 "fastdeploy/vision/common/processors/letter_box.h"
+
+namespace fastdeploy{
+namespace vision{
+bool LetterBoxResize::ImplByOpenCV(Mat* mat) {
+
+  if (mat->Channels() != color_.size()) {
+    FDERROR << "Pad: Require input channels equals to size of padding value, "
+               "but now channels = "
+            << mat->Channels()
+            << ", the size of padding values = " << color_.size() << "."
+            << std::endl;
+    return false;
+  }
+  cv::Mat* im = mat->GetOpenCVMat();
+  // generate scale_factor
+  int origin_w = mat->Width();
+  int origin_h = mat->Height();
+  int target_h = target_size_[0];
+  int target_w = target_size_[1];
+  float ratio_h = static_cast<float>(target_h) / static_cast<float>(origin_h);
+  float ratio_w = static_cast<float>(target_w) / static_cast<float>(origin_w);
+  float resize_scale = std::min(ratio_h, ratio_w);
+  // get_resized_shape
+  int new_shape_w = std::round(im->cols * resize_scale);
+  int new_shape_h = std::round(im->rows * resize_scale);
+  // calculate pad
+  float padw = (target_size_[1] - new_shape_w) / 2.;
+  float padh = (target_size_[0] - new_shape_h) / 2.;
+  int top = std::round(padh - 0.1);
+  int bottom = std::round(padh + 0.1);
+  int left = std::round(padw - 0.1);
+  int right = std::round(padw + 0.1);
+  cv::resize(*im, *im, cv::Size(new_shape_w, new_shape_h), 0, 0, cv::INTER_AREA);
+  cv::Scalar color;
+  if (color_.size() == 1) {
+      color = cv::Scalar(color_[0]);
+  } else if (color_.size() == 2) {
+      color = cv::Scalar(color_[0], color_[1]);
+  } else if (color_.size() == 3) {
+      color = cv::Scalar(color_[0], color_[1], color_[2]);
+  } else {
+      color = cv::Scalar(color_[0], color_[1], color_[2], color_[3]);
+  }
+  cv::copyMakeBorder(*im, *im, top, bottom, left, right, cv::BORDER_CONSTANT, color);
+  mat->SetWidth(im->cols);
+  mat->SetHeight(im->rows);
+  return true;
+}
+
+bool LetterBoxResize::Run(Mat* mat, const std::vector<int>& target_size, const std::vector<float>& color, ProcLib lib) {
+    auto l = LetterBoxResize(target_size,color);
+    return l(mat, lib);
+}
+
+}  // namespace vision
+}  // namespace fastdeploy

fastdeploy/vision/tracking/pptracking/letter_box.h → fastdeploy/vision/common/processors/letter_box.h

@@ -1,4 +1,4 @@
-//   Copyright (c) 2022 PaddlePaddle Authors. All Rights Reserved.
+// Copyright (c) 2022 PaddlePaddle Authors. 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.
@@ -14,23 +14,28 @@
 
 #pragma once
 
-#include "fastdeploy/vision/common/processors/transform.h"
-#include "fastdeploy/fastdeploy_model.h"
+#include "fastdeploy/vision/common/processors/base.h"
+
 namespace fastdeploy {
 namespace vision {
-namespace tracking {
-class LetterBoxResize:  public Processor{
-public:
-  LetterBoxResize(const std::vector<int>& target_size, const std::vector<float>& color);
-  bool ImplByOpenCV(Mat* mat) override;
-  std::string Name() override { return "LetterBoxResize"; }
-private:
-  std::vector<int> target_size_;
-  std::vector<float> color_;
-};
 
-} // namespace tracking
-} // namespace vision
-} // namespace fastdeploy
+class LetterBoxResize : public Processor {
+ public:
+  LetterBoxResize(const std::vector<int>& target_size, const std::vector<float>& color) {
+      target_size_ = target_size;
+      color_ = color;
+  }
+
+  bool ImplByOpenCV(Mat* mat);
 
+  std::string Name() { return "LetterBoxResize"; }
 
+  static bool Run(Mat* mat,const std::vector<int>& target_size, const std::vector<float>& color,
+                  ProcLib lib = ProcLib::OPENCV);
+
+ private:
+   std::vector<int> target_size_;
+   std::vector<float> color_;
+};
+}  // namespace vision
+}  // namespace fastdeploy

fastdeploy/vision/common/processors/transform.h

@@ -31,3 +31,4 @@
 #include "fastdeploy/vision/common/processors/resize_by_short.h"
 #include "fastdeploy/vision/common/processors/stride_pad.h"
 #include "fastdeploy/vision/common/processors/warp_affine.h"
+#include "fastdeploy/vision/common/processors/letter_box.h"

fastdeploy/vision/common/result.h

@@ -155,16 +155,25 @@ struct FASTDEPLOY_DECL OCRResult : public BaseResult {
   std::string Str();
 };
 
+/*! @brief MOT(Multi-Object Tracking) result structure for all the MOT models
+ */
 struct FASTDEPLOY_DECL MOTResult : public BaseResult {
-  // left top right bottom
+  /** \brief All the tracking object boxes for an input image, the size of `boxes` is the number of tracking objects, and the element of `boxes` is a array of 4 float values, means [xmin, ymin, xmax, ymax]
+   */
   std::vector<std::array<int, 4>> boxes;
+  /** \brief All the tracking object ids
+   */
   std::vector<int> ids;
+  /** \brief The confidence for all the tracking objects
+   */
   std::vector<float> scores;
+  /** \brief The classify label id for all the tracking object
+   */
   std::vector<int> class_ids;
   ResultType type = ResultType::MOT;
-
+  /// Clear MOT result
   void Clear();
-
+  /// Debug function, convert the result to string to print
   std::string Str();
 };
 

fastdeploy/vision/tracking/pptracking/model.cc

@@ -151,32 +151,7 @@ bool PPTracking::BuildPreprocessPipelineFromConfig(){
   return true;
 }
 
-void PPTracking::GetNmsInfo() {
-  if (runtime_option.model_format == ModelFormat::PADDLE) {
-    std::string contents;
-    if (!ReadBinaryFromFile(runtime_option.model_file, &contents)) {
-        return;
-    }
-    auto reader = paddle2onnx::PaddleReader(contents.c_str(), contents.size());
-    if (reader.has_nms) {
-      has_nms_ = true;
-      background_label = reader.nms_params.background_label;
-      keep_top_k = reader.nms_params.keep_top_k;
-      nms_eta = reader.nms_params.nms_eta;
-      nms_threshold = reader.nms_params.nms_threshold;
-      score_threshold = reader.nms_params.score_threshold;
-      nms_top_k = reader.nms_params.nms_top_k;
-      normalized = reader.nms_params.normalized;
-    }
-  }
-}
-
 bool PPTracking::Initialize() {
-  // remove multiclass_nms3 now
-  // this is a trick operation for ppyoloe while inference on trt
-  GetNmsInfo();
-  runtime_option.remove_multiclass_nms_ = true;
-  runtime_option.custom_op_info_["multiclass_nms3"] = "MultiClassNMS";
   if (!BuildPreprocessPipelineFromConfig()) {
     FDERROR << "Failed to build preprocess pipeline from configuration file."
             << std::endl;
@@ -186,18 +161,6 @@ bool PPTracking::Initialize() {
     FDERROR << "Failed to initialize fastdeploy backend." << std::endl;
     return false;
   }
-
-  if (has_nms_ && runtime_option.backend == Backend::TRT) {
-    FDINFO << "Detected operator multiclass_nms3 in your model, will replace "
-              "it with fastdeploy::backend::MultiClassNMS(background_label="
-           << background_label << ", keep_top_k=" << keep_top_k
-           << ", nms_eta=" << nms_eta << ", nms_threshold=" << nms_threshold
-           << ", score_threshold=" << score_threshold
-           << ", nms_top_k=" << nms_top_k << ", normalized=" << normalized
-           << ")." << std::endl;
-    has_nms_ = false;
-  }
-
   // create JDETracker instance
   std::unique_ptr<JDETracker> jdeTracker(new JDETracker);
   jdeTracker_ = std::move(jdeTracker);

fastdeploy/vision/tracking/pptracking/model.h

@@ -18,7 +18,7 @@
 #include "fastdeploy/fastdeploy_model.h"
 #include "fastdeploy/vision/common/result.h"
 #include "fastdeploy/vision/tracking/pptracking/tracker.h"
-#include "fastdeploy/vision/tracking/pptracking/letter_box.h"
+//#include "fastdeploy/vision/tracking/pptracking/letter_box.h"
 
 namespace fastdeploy {
 namespace vision {
@@ -56,8 +56,8 @@ class FASTDEPLOY_DECL PPTracking: public FastDeployModel {
 private:
 
   bool BuildPreprocessPipelineFromConfig();
+
   bool Initialize();
-  void GetNmsInfo();
 
   bool Preprocess(Mat* img, std::vector<FDTensor>* outputs);
 
@@ -69,19 +69,8 @@ class FASTDEPLOY_DECL PPTracking: public FastDeployModel {
   float conf_thresh_;
   float tracked_thresh_;
   float min_box_area_;
-  bool is_scale_ = true;
   std::unique_ptr<JDETracker> jdeTracker_;
 
-  // configuration for nms
-  int64_t background_label = -1;
-  int64_t keep_top_k = 300;
-  float nms_eta = 1.0;
-  float nms_threshold = 0.7;
-  float score_threshold = 0.01;
-  int64_t nms_top_k = 10000;
-  bool normalized = true;
-  bool has_nms_ = true;
-
 };
 
 } // namespace tracking

python/fastdeploy/vision/tracking/pptracking/__init__.py

@@ -24,6 +24,14 @@ def __init__(self,
                  config_file,
                  runtime_option=None,
                  model_format=ModelFormat.PADDLE):
+        """Load a PPTracking model exported by PaddleDetection.
+
+        :param model_file: (str)Path of model file, e.g pptracking/model.pdmodel
+        :param params_file: (str)Path of parameters file, e.g ppyoloe/model.pdiparams
+        :param config_file: (str)Path of configuration file for deployment, e.g ppyoloe/infer_cfg.yml
+        :param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU
+        :param model_format: (fastdeploy.ModelForamt)Model format of the loaded model
+        """
         super(PPTracking, self).__init__(runtime_option)
 
         assert model_format == ModelFormat.PADDLE, "PPTracking model only support model format of ModelFormat.Paddle now."
@@ -33,5 +41,10 @@ def __init__(self,
         assert self.initialized, "PPTracking model initialize failed."
 
     def predict(self, input_image):
+        """Predict the MOT result for an input image
+
+        :param input_image: (numpy.ndarray)The input image data, 3-D array with layout HWC, BGR format
+        :return: MOTResult
+        """
         assert input_image is not None, "The input image data is None."
         return self._model.predict(input_image)