编译paddleseg infer.cs,没推理结果

环境

• 【FastDeploy版本】： 说明具体的版本，如fastdeploy-linux-gpu-1.0.4
• 【编译命令】自行编译C# API
• cmake .. -G "Visual Studio 16 2019" -A x64 -DENABLE_ORT_BACKEND=ON -DENABLE_PADDLE_BACKEND=ON -DENABLE_OPENVINO_BACKEND=ON -DENABLE_VISION=ON -DWITH_GPU=ON -DENABLE_TEXT=ON -DWITH_CAPI=ON -DWITH_CSHARPAPI=ON -DCMAKE_INSTALL_PREFIX="D:\Paddle\compiled_fastdeploy_CUDA_csharp" -DCUDA_DIRECTORY="C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.2" -DTRT_DIRECTORY="E:\work\tools\TensorRT-8.5.2.2.Windows10.x86_64.cuda-11.8.cudnn8.6\TensorRT-8.5.2.2"
• 【系统平台】: Windows x64(Windows10)
• 【硬件】： 说明具体硬件型号，如 Nvidia GPU 3060， CUDA 11.2 CUDNN 8.4
• 【编译语言】： C#

修改了OpenCvSharp版本4.7.0.20230115为4.4.2.0.20200108，以对应net4.0版本，编译成功C# API

编译E:\work\tools\fastdeploy-win-x64-gpu-1.0.4\examples\vision\segmentation\paddleseg\cpu-gpu\csharp中的infer.cs 使用以下方法：

1、vs2019 创建控制台应用.net framework新项目D:\paddleseg_csharp\paddleseg_demo_gpu，选择net4.0版本
将E:\work\tools\fastdeploy-win-x64-gpu-1.0.4\examples\vision\segmentation\paddleseg\cpu-gpu\csharp中的infer.cs代码拷贝到D:\paddleseg_csharp\paddleseg_demo_gpu infer.cs
2、下载opencvsharp4.2.0.20200108版本，和编译fastdeploy_csharp API一致，添加引用，引用编译好的fastdeploy_csharp
3、设置配置管理器：release x64,生成paddle_demo.exe.
4、将所有dll拷贝到可执行程序同目录，fastdeploy_init.bat install %cd% D:\paddleseg_csharp\paddleseg_demo_gpu\bin\x64\Release
5、运行paddleseg_demo_gpu PP_LiteSeg_B_STDC2_cityscapes_without_argmax_infer cityscapes_demo.png 0

执行没报错，但没显示推理结果；改代码为Cv2.ImWrite("result.jpg", res_img);也未保存结果

Start with examples/vision/segmentation/paddleseg/cpu-gpu/csharp/infer.cs and reproduce the reported command using PP_LiteSeg_B_STDC2_cityscapes_without_argmax_infer and cityscapes_demo.png. Compare the execution path with the reported Cv2.ImWrite("result.jpg", res_img) attempt; the issue is resolved when the segmentation result is visibly produced or saved.