Azure.AI.DocumentIntelligence (.NET)
使用预构建和自定义模型从文档中提取文本、表格和结构化数据。
安装
dotnet add package Azure.AI.DocumentIntelligence
dotnet add package Azure.Identity
当前版本:v1.0.0(正式发布版)
环境变量
DOCUMENT_INTELLIGENCE_ENDPOINT=https://<资源名称>.cognitiveservices.azure.com/
DOCUMENT_INTELLIGENCE_API_KEY=<你的 API 密钥>
BLOB_CONTAINER_SAS_URL=https://<存储>.blob.core.windows.net/<容器>?<SAS 令牌>
AZURE_TOKEN_CREDENTIALS=prod
身份验证
Microsoft Entra 令牌凭据
using Azure.Identity;
using Azure.AI.DocumentIntelligence;
string endpoint = Environment.GetEnvironmentVariable("DOCUMENT_INTELLIGENCE_ENDPOINT");
var credential = new DefaultAzureCredential(
DefaultAzureCredential.DefaultEnvironmentVariableName
);
var client = new DocumentIntelligenceClient(new Uri(endpoint), credential);
注意:Entra ID 需要一个 自定义子域(例如 https://<资源名称>.cognitiveservices.azure.com/),而不是区域终结点。
API 密钥
string endpoint = Environment.GetEnvironmentVariable("DOCUMENT_INTELLIGENCE_ENDPOINT");
string apiKey = Environment.GetEnvironmentVariable("DOCUMENT_INTELLIGENCE_API_KEY");
var client = new DocumentIntelligenceClient(new Uri(endpoint), new AzureKeyCredential(apiKey));
客户端类型
| 客户端 | 用途 |
|---|
DocumentIntelligenceClient | 分析文档、分类文档 |
DocumentIntelligenceAdministrationClient | 构建/管理自定义模型和分类器 |
预构建模型
| 模型 ID | 描述 |
|---|
prebuilt-read | 提取文本、语言、手写内容 |
prebuilt-layout | 提取文本、表格、选择标记、结构 |
prebuilt-invoice | 提取发票字段(供应商、项目、总金额) |
prebuilt-receipt | 提取收据字段(商家、项目、总金额) |
prebuilt-idDocument | 提取身份证件字段(姓名、出生日期、地址) |
prebuilt-businessCard | 提取名片字段 |
prebuilt-tax.us.w2 | 提取 W-2 税务表单字段 |
prebuilt-healthInsuranceCard.us | 提取健康保险卡字段 |
核心工作流
1. 分析发票
using Azure.AI.DocumentIntelligence;
Uri invoiceUri = new Uri("https://example.com/invoice.pdf");
Operation<AnalyzeResult> operation = await client.AnalyzeDocumentAsync(
WaitUntil.Completed,
"prebuilt-invoice",
invoiceUri);
AnalyzeResult result = operation.Value;
foreach (AnalyzedDocument document in result.Documents)
{
if (document.Fields.TryGetValue("VendorName", out DocumentField vendorNameField)
&& vendorNameField.FieldType == DocumentFieldType.String)
{
string vendorName = vendorNameField.ValueString;
Console.WriteLine($"供应商名称:'{vendorName}',置信度:{vendorNameField.Confidence}");
}
if (document.Fields.TryGetValue("InvoiceTotal", out DocumentField invoiceTotalField)
&& invoiceTotalField.FieldType == DocumentFieldType.Currency)
{
CurrencyValue invoiceTotal = invoiceTotalField.ValueCurrency;
Console.WriteLine($"发票总金额:'{invoiceTotal.CurrencySymbol}{invoiceTotal.Amount}'");
}
if (document.Fields.TryGetValue("Items", out DocumentField itemsField)
&& itemsField.FieldType == DocumentFieldType.List)
{
foreach (DocumentField item in itemsField.ValueList)
{
var itemFields = item.ValueDictionary;
if (itemFields.TryGetValue("Description", out DocumentField descField))
Console.WriteLine($" 项目:{descField.ValueString}");
}
}
}
2. 提取布局(文本、表格、结构)
Uri fileUri = new Uri("https://example.com/document.pdf");
Operation<AnalyzeResult> operation = await client.AnalyzeDocumentAsync(
WaitUntil.Completed,
"prebuilt-layout",
fileUri);
AnalyzeResult result = operation.Value;
foreach (DocumentPage page in result.Pages)
{
Console.WriteLine($"第 {page.PageNumber} 页:{page.Lines.Count} 行,{page.Words.Count} 个单词");
foreach (DocumentLine line in page.Lines)
{
Console.WriteLine($" 行:'{line.Content}'");
}
}
foreach (DocumentTable table in result.Tables)
{
Console.WriteLine($"表格:{table.RowCount} 行 x {table.ColumnCount} 列");
foreach (DocumentTableCell cell in table.Cells)
{
Console.WriteLine($" 单元格({cell.RowIndex},{cell.ColumnIndex}):{cell.Content}");
}
}
3. 分析收据
Operation<AnalyzeResult> operation = await client.AnalyzeDocumentAsync(
WaitUntil.Completed,
"prebuilt-receipt",
receiptUri);
AnalyzeResult result = operation.Value;
foreach (AnalyzedDocument document in result.Documents)
{
if (document.Fields.TryGetValue("MerchantName", out DocumentField merchantField))
Console.WriteLine($"商家:{merchantField.ValueString}");
if (document.Fields.TryGetValue("Total", out DocumentField totalField))
Console.WriteLine($"总金额:{totalField.ValueCurrency.Amount}");
if (document.Fields.TryGetValue("TransactionDate", out DocumentField dateField))
Console.WriteLine($"日期:{dateField.ValueDate}");
}
4. 构建自定义模型
var adminClient = new DocumentIntelligenceAdministrationClient(
new Uri(endpoint),
new AzureKeyCredential(apiKey));
string modelId = "my-custom-model";
Uri blobContainerUri = new Uri("<blob-container-sas-url>");
var blobSource = new BlobContentSource(blobContainerUri);
var options = new BuildDocumentModelOptions(modelId, DocumentBuildMode.Template, blobSource);
Operation<DocumentModelDetails> operation = await adminClient.BuildDocumentModelAsync(
WaitUntil.Completed,
options);
DocumentModelDetails model = operation.Value;
Console.WriteLine($"模型 ID:{model.ModelId}");
Console.WriteLine($"创建时间:{model.CreatedOn}");
foreach (var docType in model.DocumentTypes)
{
Console.WriteLine($"文档类型:{docType.Key}");
foreach (var field in docType.Value.FieldSchema)
{
Console.WriteLine($" 字段:{field.Key},置信度:{docType.Value.FieldConfidence[field.Key]}");
}
}
5. 构建文档分类器
string classifierId = "my-classifier";
Uri blobContainerUri = new Uri("<blob-container-sas-url>");
var sourceA = new BlobContentSource(blobContainerUri) { Prefix = "TypeA/train" };
var sourceB = new BlobContentSource(blobContainerUri) { Prefix = "TypeB/train" };
var docTypes = new Dictionary<string, ClassifierDocumentTypeDetails>()
{
{ "TypeA", new ClassifierDocumentTypeDetails(sourceA) },
{ "TypeB", new ClassifierDocumentTypeDetails(sourceB) }
};
var options = new BuildClassifierOptions(classifierId, docTypes);
Operation<DocumentClassifierDetails> operation = await adminClient.BuildClassifierAsync(
WaitUntil.Completed,
options);
DocumentClassifierDetails classifier = operation.Value;
Console.WriteLine($"分类器 ID:{classifier.ClassifierId}");
6. 分类文档
string classifierId = "my-classifier";
Uri documentUri = new Uri("https://example.com/document.pdf");
var options = new ClassifyDocumentOptions(classifierId, documentUri);
Operation<AnalyzeResult> operation = await client.ClassifyDocumentAsync(
WaitUntil.Completed,
options);
AnalyzeResult result = operation.Value;
foreach (AnalyzedDocument document in result.Documents)
{
Console.WriteLine($"文档类型:{document.DocumentType},置信度:{document.Confidence}");
}
7. 管理模型
DocumentIntelligenceResourceDetails resourceDetails = await adminClient.GetResourceDetailsAsync();
Console.WriteLine($"自定义模型:{resourceDetails.CustomDocumentModels.Count}/{resourceDetails.CustomDocumentModels.Limit}");
DocumentModelDetails model = await adminClient.GetModelAsync("my-model-id");
Console.WriteLine($"模型:{model.ModelId},创建时间:{model.CreatedOn}");
await foreach (DocumentModelDetails modelItem in adminClient.GetModelsAsync())
{
Console.WriteLine($"模型:{modelItem.ModelId}");
}
await adminClient.DeleteModelAsync("my-model-id");
关键类型参考
| 类型 | 描述 |
|---|
DocumentIntelligenceClient | 用于分析的主客户端 |
DocumentIntelligenceAdministrationClient | 模型管理 |
AnalyzeResult | 文档分析的结果 |
AnalyzedDocument | 结果中的单个文档 |
DocumentField | 提取的字段,包含值和置信度 |
DocumentFieldType | 字符串、日期、数字、货币等 |
DocumentPage | 页面信息(行、单词、选择标记) |
DocumentTable | 提取的表格,包含单元格 |
DocumentModelDetails | 自定义模型元数据 |
BlobContentSource | 训练数据源 |
构建模式
| 模式 | 使用场景 |
|---|
DocumentBuildMode.Template | 固定布局文档(表单) |
DocumentBuildMode.Neural | 可变布局文档 |
最佳实践
- 生产环境使用 DefaultAzureCredential
- 重用客户端实例 — 客户端是线程安全的
- 处理长时间运行的操作 — 为简单起见使用
WaitUntil.Completed
- 检查字段置信度 — 始终验证
Confidence 属性
- 使用合适的模型 — 常见文档用预构建,专用文档用自定义
- 使用自定义子域 — Entra ID 身份验证所必需
错误处理
using Azure;
try
{
var operation = await client.AnalyzeDocumentAsync(
WaitUntil.Completed,
"prebuilt-invoice",
documentUri);
}
catch (RequestFailedException ex)
{
Console.WriteLine($"错误:{ex.Status} - {ex.Message}");
}
相关 SDK
| SDK | 用途 | 安装 |
|---|
Azure.AI.DocumentIntelligence | 文档分析(此 SDK) | dotnet add package Azure.AI.DocumentIntelligence |
Azure.AI.FormRecognizer | 旧版 SDK(已弃用) | 改用 DocumentIntelligence |
参考链接