[Model] Support Paddle3D PETR v2 model (#1863)

* Support PETR v2

* make petrv2 precision equal with the origin repo

* delete extra func

* modify review problem

* delete visualize

* Update README_CN.md

* Update README.md

* Update README_CN.md

* fix build problem

* delete external variable and function

---------

Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com>

examples/vision/perception/paddle3d/petr/python/README.md

@@ -0,0 +1,63 @@
+English | [简体中文](README_CN.md)
+# Petr Python Deployment Example
+
+Before deployment, the following two steps need to be confirmed
+
+- 1. The hardware and software environment meets the requirements, refer to [FastDeploy environment requirements](../../../../../docs/en/build_and_install/download_prebuilt_libraries.md)
+- 2. FastDeploy Python whl package installation, refer to [FastDeploy Python Installation](../../../../../docs/cn/build_and_install/download_prebuilt_libraries.md)
+
+This directory provides an example of `infer.py` to quickly complete the deployment of Petr on CPU/GPU. Execute the following script to complete
+
+```bash
+#Download deployment sample code
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd examples/vision/vision/paddle3d/petr/python
+
+wget https://bj.bcebos.com/fastdeploy/models/petr.tar.gz
+tar -xf petr.tar.gz
+wget https://bj.bcebos.com/fastdeploy/models/petr_test.png
+
+# CPU reasoning
+python infer.py --model petr --image petr_test.png --device cpu
+# GPU inference
+python infer.py --model petr --image petr_test.png --device gpu
+```
+
+## Petr Python interface
+
+```python
+fastdeploy.vision.detection.Petr(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+```
+
+Petr model loading and initialization.
+
+**parameter**
+> * **model_file**(str): model file path
+> * **params_file**(str): parameter file path
+> * **config_file**(str): configuration file path
+> * **runtime_option**(RuntimeOption): Backend reasoning configuration, the default is None, that is, the default configuration is used
+> * **model_format**(ModelFormat): model format, the default is Paddle format
+
+### predict function
+
+> ```python
+> Petr. predict(image_data)
+> ```
+>
+> Model prediction interface, the input image directly outputs the detection result.
+>
+> **parameters**
+>
+> > * **image_data**(np.ndarray): input data, note that it must be in HWC, BGR format
+
+> **Back**
+>
+> > Return the `fastdeploy.vision.PerceptionResult` structure, structure description reference document [Vision Model Prediction Results](../../../../../docs/api/vision_results/)
+
+
+## Other documents
+
+- [Petr Model Introduction](..)
+- [Petr C++ deployment](../cpp)
+- [Description of model prediction results](../../../../../docs/api/vision_results/)
+- [How to switch model inference backend engine](../../../../../docs/en/faq/how_to_change_backend.md)

examples/vision/perception/paddle3d/petr/python/README_CN.md

@@ -0,0 +1,65 @@
+[English](README.md) | 简体中文
+
+# Petr Python 部署示例
+
+在部署前，需确认以下两个步骤
+
+- 1. 软硬件环境满足要求，参考[FastDeploy环境要求](../../../../../docs/cn/build_and_install/download_prebuilt_libraries.md)  
+- 2. FastDeploy Python whl 包安装，参考[FastDeploy Python安装](../../../../../docs/cn/build_and_install/download_prebuilt_libraries.md)
+
+本目录下提供 `infer.py` 快速完成 Petr 在 CPU/GPU上部署的示例。执行如下脚本即可完成
+
+```bash
+#下载部署示例代码
+git clone https://github.com/PaddlePaddle/FastDeploy.git
+cd examples/vision/vision/paddle3d/petr/python
+
+wget https://bj.bcebos.com/fastdeploy/models/petr.tar.gz
+tar -xf petr.tar.gz
+wget https://bj.bcebos.com/fastdeploy/models/petr_test.png
+
+# CPU推理
+python infer.py --model petr --image petr_test.png --device cpu
+# GPU推理
+python infer.py --model petr --image petr_test.png --device gpu
+```
+
+
+## Petr Python接口
+
+```python
+fastdeploy.vision.perception.Petr(model_file, params_file, config_file, runtime_option=None, model_format=ModelFormat.PADDLE)
+```
+
+Petr模型加载和初始化。
+
+**参数**
+> * **model_file**(str): 模型文件路径
+> * **params_file**(str): 参数文件路径
+> * **config_file**(str): 配置文件路径
+> * **runtime_option**(RuntimeOption): 后端推理配置，默认为None，即采用默认配置
+> * **model_format**(ModelFormat): 模型格式，默认为Paddle格式
+
+### predict 函数
+
+> ```python
+> Petr.predict(image_data)
+> ```
+>
+> 模型预测结口，输入图像直接输出检测结果。
+>
+> **参数**
+>
+> > * **image_data**(np.ndarray): 输入数据，注意需为HWC，BGR格式
+
+> **返回**
+>
+> > 返回`fastdeploy.vision.PerceptionResult`结构体，结构体说明参考文档[视觉模型预测结果](../../../../../docs/api/vision_results/)
+
+
+## 其它文档
+
+- [Petr 模型介绍](..)
+- [Petr C++部署](../cpp)
+- [模型预测结果说明](../../../../../docs/api/vision_results/)
+- [如何切换模型推理后端引擎](../../../../../docs/cn/faq/how_to_change_backend.md)

examples/vision/perception/paddle3d/petr/python/infer.py

@@ -0,0 +1,45 @@
+import fastdeploy as fd
+import cv2
+import os
+from fastdeploy import ModelFormat
+
+
+def parse_arguments():
+    import argparse
+    import ast
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--model", required=True, help="Path of petr paddle model.")
+    parser.add_argument(
+        "--image", required=True, help="Path of test image file.")
+    parser.add_argument(
+        "--device",
+        type=str,
+        default='cpu',
+        help="Type of inference device, support 'cpu' or 'gpu'.")
+    return parser.parse_args()
+
+
+def build_option(args):
+    option = fd.RuntimeOption()
+    if args.device.lower() == "gpu":
+        option.use_gpu(0)
+    if args.device.lower() == "cpu":
+        option.use_cpu()
+    return option
+
+
+args = parse_arguments()
+
+model_file = os.path.join(args.model, "petrv2_inference.pdmodel")
+params_file = os.path.join(args.model, "petrv2_inference.pdiparams")
+config_file = os.path.join(args.model, "infer_cfg.yml")
+# 配置runtime，加载模型
+runtime_option = build_option(args)
+model = fd.vision.perception.Petr(
+    model_file, params_file, config_file, runtime_option=runtime_option)
+
+# 预测图片检测结果
+im = cv2.imread(args.image)
+result = model.predict(im)
+print(result)