[FastDeploy] Improve fastdeploy code and readme (#9379)

* Fix padding value in rec model, and box sort in det model

* Add FastDeploy support to deploy PaddleOCR models.

* Improve readme

* improve readme

* improve readme

* improve code

deploy/fastdeploy/cpu-gpu/cpp/infer.cc

@@ -152,9 +152,9 @@ int main(int argc, char *argv[]) {
     option.UsePaddleBackend(); // Paddle Inference
   } else if (flag == 5) {
     option.UseGpu();
-    option.UseTrtBackend();
-    option.EnablePaddleTrtCollectShape();
-    option.EnablePaddleToTrt(); // Paddle-TensorRT
+    option.UsePaddleInferBackend();
+    option.paddle_infer_option.collect_trt_shape = true;
+    option.paddle_infer_option.enable_trt = true; // Paddle-TensorRT
   } else if (flag == 6) {
     option.UseGpu();
     option.UseOrtBackend(); // ONNX Runtime

deploy/fastdeploy/cpu-gpu/csharp/README.md

@@ -70,7 +70,7 @@ infer_demo ./ch_PP-OCRv3_det_infer ./ch_ppocr_mobile_v3.0_cls_infer ./ch_PP-OCRv
 ## 5. PP-OCRv3 C# API接口简介
 下面提供了PP-OCRv3的C# API简介
 
-- 如果用户想要更换部署后端或进行其他定制化操作, 请查看[C# Runtime API](https://github.com/PaddlePaddle/FastDeploy/blob/develop/csharp/fastdeploy/runtime_option.cs).
+- 如果用户想要更换部署后端或进行其他定制化操作, 请查看[C# Runtime API](https://baidu-paddle.github.io/fastdeploy-api/csharp/html/classfastdeploy_1_1RuntimeOption.html).
 - 更多 PP-OCR C# API 请查看 [C# PP-OCR API](https://github.com/PaddlePaddle/FastDeploy/blob/develop/csharp/fastdeploy/vision/ocr/model.cs)
 
 ### 模型

deploy/fastdeploy/cpu-gpu/python/infer.py

@@ -105,17 +105,17 @@ def build_option(args):
     elif args.backend.lower() == "pptrt":
         assert args.device.lower(
         ) == "gpu", "Paddle-TensorRT backend require inference on device GPU."
-        det_option.use_trt_backend()
-        det_option.enable_paddle_trt_collect_shape()
-        det_option.enable_paddle_to_trt()
+        det_option.use_paddle_infer_backend()
+        det_option.paddle_infer_option.collect_trt_shape = True
+        det_option.paddle_infer_option.enable_trt = True
 
-        cls_option.use_trt_backend()
-        cls_option.enable_paddle_trt_collect_shape()
-        cls_option.enable_paddle_to_trt()
+        cls_option.use_paddle_infer_backend()
+        cls_option.paddle_infer_option.collect_trt_shape = True
+        cls_option.paddle_infer_option.enable_trt = True
 
-        rec_option.use_trt_backend()
-        rec_option.enable_paddle_trt_collect_shape()
-        rec_option.enable_paddle_to_trt()
+        rec_option.use_paddle_infer_backend()
+        rec_option.paddle_infer_option.collect_trt_shape = True
+        rec_option.paddle_infer_option.enable_trt = True
 
         # If use TRT backend, the dynamic shape will be set as follow.
         # We recommend that users set the length and height of the detection model to a multiple of 32.