升级多个依赖&增加额外的推理加速功能&迁移OCR (#1459)

* 更新多个NuGet包至最新版本

* 重构模型加载以适应yolosharp

* feat: 改变接口。TensorRT缓存的初步支持，修改配置项。

* 更新依赖并调整变量顺序，修复加载问题

* 更新AvalonEdit和Microsoft.ML.OnnxRuntime包至最新版本,以修复问题

* fix: downgrade Microsoft.ML.OnnxRuntime.DirectML to version 1.21.0

* typo

* fix: change log level from warning to error for ONNX provider loading failure

* 增加 paddle ocr 的 onnx 模型

* feat: add PaddleOCR models for Chinese, English, and Latin recognition

* 使用cv的DNN生成Tensor，加速Yap文字识别

* feat: 尝试搓一个onnx的ocr

* clean up code

* chore: update OpenCvSharp4 package versions to 4.10.0.20241108

* 修复因格式化代码而丢的引用

* chore: update Microsoft.ML.OnnxRuntime.DirectML package to version 1.21.1 and improve logging for ONNX provider initialization

* chore: 等yolosharp更新再升级onnx

* chore: add Microsoft.ML.OnnxRuntime.Managed package and clean up logging in Det class

* fix: refactor output tensor handling in Det class for improved clarity

* 补充注释，修复DML的OCR问题

* 默认OCR推理使用CPU,整理配置

* fix error NETSDK1152: 找到了多个具有相同相对路径的发布输出文件

* fix(logging): enhance debug log for ONNX initialization with provider details

* 修复TensorRT模型缓存的加载问题

* fix(onnx): improve cached model retrieval and add file existence check

* fix(ocr): replace SrcGreyMat with SrcMat for region of interest processing

* fix(onnx): add file existence check for cached model and adjust session options for DirectML provider

* 增加硬件加速配置UI界面

* 移除旧的OCR模型

* 错别字

---------

Co-authored-by: 辉鸭蛋 <huiyadanli@gmail.com>

Author: Takaranoao