Please use. NET 5,. NET 6, NET 7,. NET 8, NET Framework 4.6, NET Framework 4.61, NET Framework 4.7, NET Framework 4.72, NET Framework 4.8, NET Framework 4.81, and. NET Core 3.1 versions

DeploySharp.Data 命名空间

Handles data processing, transformation and preparation tasks 负责数据处理、转换和准备工作

备注

Contains functionality for data preprocessing, postprocessing, normalization, and conversion between different data formats. 包含数据预处理、后处理、标准化及不同数据格式之间转换的功能

AnomalySegResult 
BoundingBox Represents a bounding box with confidence score, spatial coordinates, and metadata 表示具有置信度分数、空间坐标和元数据的边界框
CvDataExtensions Provides extension methods for converting between SixLabors.ImageSharp and DeploySharp CVData data structures 提供SixLabors.ImageSharp和DeploySharp CVData数据结构之间的转换扩展方法
CvDataProcessor Provides image processing utilities for computer vision tasks 提供计算机视觉任务的图像处理工具
DataProcessorConfig Configuration settings for data processing pipeline 数据预处理管线的配置设置
DataTensor Manages collection of tensor nodes with validation and enumeration capabilities 管理具有验证和枚举功能的张量节点集合
DetResult Represents object detection results with bounding box information 表示包含边界框信息的物体检测结果
ImageDataT Generic image container supporting both CHW (channel-height-width) and HWC (height-width-channel) layouts 支持CHW(通道-高度-宽度)和HWC(高度-宽度-通道)布局的通用图像容器
ImageDataB Specialized byte-based image container with additional image processing operations 专门针对字节数据的图像容器,提供额外的图像处理操作
ImageDataF Specialized floating-point image container optimized for machine learning operations 专为机器学习优化的浮点图像容器
KeyPointResult Represents a collection of keypoints detected in an image 表示图像中检测到的关键点集合
NodeData Represents a single tensor node with metadata and data buffer 表示具有元数据和数据缓冲区的单个张量节点
NonMaxSuppression Performs Non-Maximum Suppression (NMS) to filter overlapping bounding boxes 执行非极大值抑制(NMS)以过滤重叠的边界框
NormalizationParams Contains configuration parameters for normalization operations 包含归一化操作的配置参数
NormalizationParamsFactory Factory class providing preset normalization parameters 提供预设归一化参数的工厂类
ObbNonMaxSuppression Oriented bounding box implementation of NMS 定向边界框(OBB)的NMS实现
ObbResult Represents object detection results with oriented bounding boxes (OBB) 表示带有方向边界框(OBB)的物体检测结果
PairT1, T2 Represents an immutable generic pair of values with multiple functional interfaces 表示具有多个功能接口的不可变泛型值对
PairExtensions Provides extension methods for Pair objects 为Pair对象提供扩展方法
RectNonMaxSuppression Axis-aligned rectangle implementation of NMS 轴对齐矩形的NMS实现
Result Represents the base result structure for model inference outputs 表示模型推理输出的基础结果结构
ResultSetResultUnit Represents a collection of prediction results with generic type support 表示具有泛型类型支持的预测结果集合
SegResult Represents semantic segmentation results including pixel-wise classification mask 表示语义分割结果,包含像素级分类掩码
TripletT1, T2, T3 Represents an immutable generic triplet of values 表示不可变泛型三元组
VisionColors Computer vision color provider (supports automatic coloring for 80 class IDs) 计算机视觉颜色提供器(支持80类别ID的自动配色)
Visualize Static class providing visualization methods for different computer vision results 提供不同计算机视觉结果可视化方法的静态类
VisualizeVisualizeHandler Handler class for polymorphic visualization operations 多态可视化操作的处理程序类
VisualizeHandler 
VisualizeOptions Configuration options for visualization rendering 可视化渲染的配置选项

结构

ImageAdjustmentParam Image adjustment parameters (internal structure) 图像调整参数(内部结构体)
KeyPoint Represents a single detected keypoint with location and confidence score 表示带有位置和置信度分数的单个检测关键点
Point Represents a 2D point with integer coordinates and geometric operations 表示具有整数坐标和几何运算的二维点
PointD Represents a high-precision 2D point with double-precision coordinates and advanced geometric operations 表示具有双精度坐标和高级几何运算的高精度二维点
PointF Represents a single-precision floating-point 2D point with geometric operations 表示具有几何运算的单精度浮点二维点
Rect Represents a rectangular region with integer coordinates and dimensions 表示具有整数坐标和尺寸的矩形区域
RectD Represents a rectangle defined by X/Y coordinates, width and height (double precision) 表示由X/Y坐标、宽度和高度定义的双精度矩形结构
RectF Represents a rectangle structure with single-precision floating-point coordinates and dimensions 表示具有单精度浮点坐标和尺寸的矩形结构
RotatedRect Represents a rectangle that can be rotated around its center point 表示可围绕其中心点旋转的矩形
Size Represents a size with integer width and height dimensions 表示具有整数宽度和高度尺寸的大小
SizeD Represents a size with double-precision floating-point dimensions 表示具有双精度浮点尺寸的大小结构体
SizeF Represents a size with single-precision floating-point dimensions 表示具有单精度浮点尺寸的大小结构体

委托

VisualizeVisualizeHandlerVisualizeDelegate Delegate type for visualization methods 可视化方法的委托类型
VisualizeHandlerVisualizeDelegate 

枚举

ImageDataTDataFormat Supported data layout formats 支持的数据布局格式
ImageNormalizationType Specifies image normalization methods 指定图像归一化处理方法
ImageResizeMode Specifies how images should be resized during preprocessing 指定预处理期间图像应该如何调整大小
ResultType Defines types of model inference results 定义模型推理结果的类型
TensorType Enumeration defining tensor types (Input/Output) 定义张量类型的枚举(输入/输出)