AWS SDK または CLI DetectDocumentTextで使用する

ドキュメント内のテキストを検出するには

次の detect-document-text の例は、ドキュメント内のテキストを検出する方法を示しています。

Linux/macOS:

aws textract detect-document-text \
    --document '{"S3Object":{"Bucket":"bucket","Name":"document"}}'

Windows ：

aws textract detect-document-text \
    --document "{\"S3Object\":{\"Bucket\":\"bucket\",\"Name\":\"document\"}}" \
    --region region-name

出力:

{
    "Blocks": [
        {
            "Geometry": {
                "BoundingBox": {
                    "Width": 1.0,
                    "Top": 0.0,
                    "Left": 0.0,
                    "Height": 1.0
                },
                "Polygon": [
                    {
                        "Y": 0.0,
                        "X": 0.0
                    },
                    {
                        "Y": 0.0,
                        "X": 1.0
                    },
                    {
                        "Y": 1.0,
                        "X": 1.0
                    },
                    {
                        "Y": 1.0,
                        "X": 0.0
                    }
                ]
            },
            "Relationships": [
                {
                    "Type": "CHILD",
                    "Ids": [
                        "896a9f10-9e70-4412-81ce-49ead73ed881",
                        "0da18623-dc4c-463d-a3d1-9ac050e9e720",
                        "167338d7-d38c-4760-91f1-79a8ec457bb2"
                    ]
                }
            ],
            "BlockType": "PAGE",
            "Id": "21f0535e-60d5-4bc7-adf2-c05dd851fa25"
        },
        {
            "Relationships": [
                {
                    "Type": "CHILD",
                    "Ids": [
                        "62490c26-37ea-49fa-8034-7a9ff9369c9c",
                        "1e4f3f21-05bd-4da9-ba10-15d01e66604c"
                    ]
                }
            ],
            "Confidence": 89.11581420898438,
            "Geometry": {
                "BoundingBox": {
                    "Width": 0.33642634749412537,
                    "Top": 0.17169663310050964,
                    "Left": 0.13885067403316498,
                    "Height": 0.49159330129623413
                },
                "Polygon": [
                    {
                        "Y": 0.17169663310050964,
                        "X": 0.13885067403316498
                    },
                    {
                        "Y": 0.17169663310050964,
                        "X": 0.47527703642845154
                    },
                    {
                        "Y": 0.6632899641990662,
                        "X": 0.47527703642845154
                    },
                    {
                        "Y": 0.6632899641990662,
                        "X": 0.13885067403316498
                    }
                ]
            },
            "Text": "He llo,",
            "BlockType": "LINE",
            "Id": "896a9f10-9e70-4412-81ce-49ead73ed881"
        },
        {
            "Relationships": [
                {
                    "Type": "CHILD",
                    "Ids": [
                        "19b28058-9516-4352-b929-64d7cef29daf"
                    ]
                }
            ],
            "Confidence": 85.5694351196289,
            "Geometry": {
                "BoundingBox": {
                    "Width": 0.33182239532470703,
                    "Top": 0.23131252825260162,
                    "Left": 0.5091826915740967,
                    "Height": 0.3766750991344452
                },
                "Polygon": [
                    {
                        "Y": 0.23131252825260162,
                        "X": 0.5091826915740967
                    },
                    {
                        "Y": 0.23131252825260162,
                        "X": 0.8410050868988037
                    },
                    {
                        "Y": 0.607987642288208,
                        "X": 0.8410050868988037
                    },
                    {
                        "Y": 0.607987642288208,
                        "X": 0.5091826915740967
                    }
                ]
            },
            "Text": "worlc",
            "BlockType": "LINE",
            "Id": "0da18623-dc4c-463d-a3d1-9ac050e9e720"
        }
    ],
    "DocumentMetadata": {
        "Pages": 1
    }
}

詳細については、「Amazon Textract デベロッパーガイド」の「ドキュメントのテキストの検出」を参照してください。

入力ドキュメントからテキストを検出します。

import software.amazon.awssdk.core.SdkBytes;
import software.amazon.awssdk.regions.Region;
import software.amazon.awssdk.services.textract.TextractClient;
import software.amazon.awssdk.services.textract.model.Document;
import software.amazon.awssdk.services.textract.model.DetectDocumentTextRequest;
import software.amazon.awssdk.services.textract.model.DetectDocumentTextResponse;
import software.amazon.awssdk.services.textract.model.Block;
import software.amazon.awssdk.services.textract.model.DocumentMetadata;
import software.amazon.awssdk.services.textract.model.TextractException;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.InputStream;
import java.util.List;

/**
 * Before running this Java V2 code example, set up your development
 * environment, including your credentials.
 *
 * For more information, see the following documentation topic:
 *
 * https://docs.aws.amazon.com/sdk-for-java/latest/developer-guide/get-started.html
 */
public class DetectDocumentText {
    public static void main(String[] args) {
        final String usage = """

                Usage:
                    <sourceDoc>\s

                Where:
                    sourceDoc - The path where the document is located (must be an image, for example, C:/AWS/book.png).\s
                """;

        if (args.length != 1) {
            System.out.println(usage);
            System.exit(1);
        }

        String sourceDoc = args[0];
        Region region = Region.US_EAST_2;
        TextractClient textractClient = TextractClient.builder()
                .region(region)
                .build();

        detectDocText(textractClient, sourceDoc);
        textractClient.close();
    }

    public static void detectDocText(TextractClient textractClient, String sourceDoc) {
        try {
            InputStream sourceStream = new FileInputStream(new File(sourceDoc));
            SdkBytes sourceBytes = SdkBytes.fromInputStream(sourceStream);

            // Get the input Document object as bytes.
            Document myDoc = Document.builder()
                    .bytes(sourceBytes)
                    .build();

            DetectDocumentTextRequest detectDocumentTextRequest = DetectDocumentTextRequest.builder()
                    .document(myDoc)
                    .build();

            // Invoke the Detect operation.
            DetectDocumentTextResponse textResponse = textractClient.detectDocumentText(detectDocumentTextRequest);
            List<Block> docInfo = textResponse.blocks();
            for (Block block : docInfo) {
                System.out.println("The block type is " + block.blockType().toString());
            }

            DocumentMetadata documentMetadata = textResponse.documentMetadata();
            System.out.println("The number of pages in the document is " + documentMetadata.pages());

        } catch (TextractException | FileNotFoundException e) {

            System.err.println(e.getMessage());
            System.exit(1);
        }
    }
}

Amazon S3 バケット内のドキュメントからテキストを検出する

import software.amazon.awssdk.regions.Region;
import software.amazon.awssdk.services.textract.model.S3Object;
import software.amazon.awssdk.services.textract.TextractClient;
import software.amazon.awssdk.services.textract.model.Document;
import software.amazon.awssdk.services.textract.model.DetectDocumentTextRequest;
import software.amazon.awssdk.services.textract.model.DetectDocumentTextResponse;
import software.amazon.awssdk.services.textract.model.Block;
import software.amazon.awssdk.services.textract.model.DocumentMetadata;
import software.amazon.awssdk.services.textract.model.TextractException;

/**
 * Before running this Java V2 code example, set up your development
 * environment, including your credentials.
 *
 * For more information, see the following documentation topic:
 *
 * https://docs.aws.amazon.com/sdk-for-java/latest/developer-guide/get-started.html
 */
public class DetectDocumentTextS3 {

    public static void main(String[] args) {
        final String usage = """

                Usage:
                    <bucketName> <docName>\s

                Where:
                    bucketName - The name of the Amazon S3 bucket that contains the document.\s

                    docName - The document name (must be an image, i.e., book.png).\s
                """;

        if (args.length != 2) {
            System.out.println(usage);
            System.exit(1);
        }

        String bucketName = args[0];
        String docName = args[1];
        Region region = Region.US_WEST_2;
        TextractClient textractClient = TextractClient.builder()
                .region(region)
                .build();

        detectDocTextS3(textractClient, bucketName, docName);
        textractClient.close();
    }

    public static void detectDocTextS3(TextractClient textractClient, String bucketName, String docName) {
        try {
            S3Object s3Object = S3Object.builder()
                    .bucket(bucketName)
                    .name(docName)
                    .build();

            // Create a Document object and reference the s3Object instance.
            Document myDoc = Document.builder()
                    .s3Object(s3Object)
                    .build();

            DetectDocumentTextRequest detectDocumentTextRequest = DetectDocumentTextRequest.builder()
                    .document(myDoc)
                    .build();

            DetectDocumentTextResponse textResponse = textractClient.detectDocumentText(detectDocumentTextRequest);
            for (Block block : textResponse.blocks()) {
                System.out.println("The block type is " + block.blockType().toString());
            }

            DocumentMetadata documentMetadata = textResponse.documentMetadata();
            System.out.println("The number of pages in the document is " + documentMetadata.pages());

        } catch (TextractException e) {

            System.err.println(e.getMessage());
            System.exit(1);
        }
    }
}

class TextractWrapper:
    """Encapsulates Textract functions."""

    def __init__(self, textract_client, s3_resource, sqs_resource):
        """
        :param textract_client: A Boto3 Textract client.
        :param s3_resource: A Boto3 Amazon S3 resource.
        :param sqs_resource: A Boto3 Amazon SQS resource.
        """
        self.textract_client = textract_client
        self.s3_resource = s3_resource
        self.sqs_resource = sqs_resource


    def detect_file_text(self, *, document_file_name=None, document_bytes=None):
        """
        Detects text elements in a local image file or from in-memory byte data.
        The image must be in PNG or JPG format.

        :param document_file_name: The name of a document image file.
        :param document_bytes: In-memory byte data of a document image.
        :return: The response from Amazon Textract, including a list of blocks
                 that describe elements detected in the image.
        """
        if document_file_name is not None:
            with open(document_file_name, "rb") as document_file:
                document_bytes = document_file.read()
        try:
            response = self.textract_client.detect_document_text(
                Document={"Bytes": document_bytes}
            )
            logger.info("Detected %s blocks.", len(response["Blocks"]))
        except ClientError:
            logger.exception("Couldn't detect text.")
            raise
        else:
            return response

    "Detects text in the input document."
    "Amazon Textract can detect lines of text and the words that make up a line of text."
    "The input document must be in one of the following image formats: JPEG, PNG, PDF, or TIFF."

    "Create an ABAP object for the Amazon S3 object."
    DATA(lo_s3object) = NEW /aws1/cl_texs3object( iv_bucket = iv_s3bucket
      iv_name   = iv_s3object ).

    "Create an ABAP object for the document."
    DATA(lo_document) = NEW /aws1/cl_texdocument( io_s3object = lo_s3object ).
    "Analyze document stored in Amazon S3."
    TRY.
        oo_result = lo_tex->detectdocumenttext( io_document = lo_document ).         "oo_result is returned for testing purposes."
        LOOP AT oo_result->get_blocks( ) INTO DATA(lo_block).
          IF lo_block->get_text( ) = 'INGREDIENTS: POWDERED SUGAR* (CANE SUGAR,'.
            MESSAGE 'Found text in the doc: ' && lo_block->get_text( ) TYPE 'I'.
          ENDIF.
        ENDLOOP.
        DATA(lo_metadata) = oo_result->get_documentmetadata( ).
        MESSAGE 'The number of pages in the document is ' && lo_metadata->ask_pages( ) TYPE 'I'.
        MESSAGE 'Detect document text completed.' TYPE 'I'.
      CATCH /aws1/cx_texaccessdeniedex.
        MESSAGE 'You do not have permission to perform this action.' TYPE 'E'.
      CATCH /aws1/cx_texbaddocumentex.
        MESSAGE 'Amazon Textract is not able to read the document.' TYPE 'E'.
      CATCH /aws1/cx_texdocumenttoolargeex.
        MESSAGE 'The document is too large.' TYPE 'E'.
      CATCH /aws1/cx_texinternalservererr.
        MESSAGE 'Internal server error.' TYPE 'E'.
      CATCH /aws1/cx_texinvalidparameterex.
        MESSAGE 'Request has non-valid parameters.' TYPE 'E'.
      CATCH /aws1/cx_texinvalids3objectex.
        MESSAGE 'Amazon S3 object is not valid.' TYPE 'E'.
      CATCH /aws1/cx_texprovthruputexcdex.
        MESSAGE 'Provisioned throughput exceeded limit.' TYPE 'E'.
      CATCH /aws1/cx_texthrottlingex.
        MESSAGE 'The request processing exceeded the limit' TYPE 'E'.
      CATCH /aws1/cx_texunsupporteddocex.
        MESSAGE 'The document is not supported.' TYPE 'E'.
    ENDTRY.