[FR] Selectionmarks

sdk/formrecognizer/Azure.AI.FormRecognizer/CHANGELOG.md

@@ -2,7 +2,8 @@
 
 ## 3.1.0-beta.1 (Unreleased)
 - It defaults to the latest supported API version, which currently is `2.1-preview.1`.
-Note that new functionality hasn't been implemented in the client library.
+- Added support to `StartRecognizeContent` to recognize selection marks such as check boxes and radio buttons.
+- Added support to train and recognize custom forms with selection marks such as check boxes and radio buttons. This functionality is only available in train with labels scenarios.
 
 ## 3.0.0 (2020-08-20)
 

sdk/formrecognizer/Azure.AI.FormRecognizer/README.md

@@ -2,7 +2,7 @@
 Azure Cognitive Services Form Recognizer is a cloud service that uses machine learning to recognize form fields, text, and tables in form documents.  It includes the following capabilities:
 
 - Recognize Custom Forms - Recognize and extract form fields and other content from your custom forms, using models you trained with your own form types.
-- Recognize Form Content - Recognize and extract tables, lines and words in forms documents, without the need to train a model.
+- Recognize Form Content - Recognize and extract tables, lines, words, and selection marks like radio buttons and check boxes in forms documents, without the need to train a model.
 - Recognize Receipts - Recognize and extract common fields from US receipts, using a pre-trained receipt model.
 
 [Source code][formreco_client_src] | [Package (NuGet)][formreco_nuget_package] | [API reference documentation][formreco_refdocs] | [Product documentation][formreco_docs] | [Samples][formreco_samples]
@@ -101,7 +101,7 @@ var client = new FormRecognizerClient(new Uri(endpoint), new DefaultAzureCredent
 `FormRecognizerClient` provides operations for:
 
  - Recognizing form fields and content, using custom models trained to recognize your custom forms.  These values are returned in a collection of `RecognizedForm` objects. See example [Recognize Custom Forms](#recognize-custom-forms).
- - Recognizing form content, including tables, lines and words, without the need to train a model.  Form content is returned in a collection of `FormPage` objects. See example [Recognize Content](#recognize-content).
+ - Recognizing form content, including tables, lines, words, and selection marks like radio buttons and check boxes without the need to train a model.  Form content is returned in a collection of `FormPage` objects. See example [Recognize Content](#recognize-content).
  - Recognizing common fields from US receipts, using a pre-trained receipt model on the Form Recognizer service.  These fields and meta-data are returned in a collection of `RecognizedForm` objects. See example [Recognize Receipts](#recognize-receipts).
 
 ### FormTrainingClient
@@ -137,7 +137,7 @@ The following section provides several code snippets illustrating common pattern
 * [Manage Custom Models Synchronously](#manage-custom-models-synchronously)
 
 ### Recognize Content
-Recognize text and table data, along with their bounding box coordinates, from documents.
+Recognize text, tables, and selection marks like radio buttons and check boxes data, along with their bounding box coordinates, from documents.
 
 ```C# Snippet:FormRecognizerSampleRecognizeContentFromUri
 FormPageCollection formPages = await client.StartRecognizeContentFromUriAsync(invoiceUri).WaitForCompletionAsync();
@@ -160,6 +160,17 @@ foreach (FormPage page in formPages)
             Console.WriteLine($"    Cell ({cell.RowIndex}, {cell.ColumnIndex}) contains text: '{cell.Text}'.");
         }
     }
+
+    for (int i = 0; i < page.SelectionMarks.Count; i++)
+    {
+        FormSelectionMark selectionMark = page.SelectionMarks[i];
+        Console.WriteLine($"Selection Mark {i} is {selectionMark.State.ToString()}.");
+        Console.WriteLine("        Its bounding box is:");
+        Console.WriteLine($"        Upper left => X: {selectionMark.BoundingBox[0].X}, Y= {selectionMark.BoundingBox[0].Y}");
+        Console.WriteLine($"        Upper right => X: {selectionMark.BoundingBox[1].X}, Y= {selectionMark.BoundingBox[1].Y}");
+        Console.WriteLine($"        Lower right => X: {selectionMark.BoundingBox[2].X}, Y= {selectionMark.BoundingBox[2].Y}");
+        Console.WriteLine($"        Lower left => X: {selectionMark.BoundingBox[3].X}, Y= {selectionMark.BoundingBox[3].Y}");
+    }
 }
 ```
 

sdk/formrecognizer/Azure.AI.FormRecognizer/api/Azure.AI.FormRecognizer.netstandard2.0.cs

@@ -88,6 +88,7 @@ public readonly partial struct FieldValue
         public System.DateTime AsDate() { throw null; }
         public System.Collections.Generic.IReadOnlyDictionary<string, Azure.AI.FormRecognizer.Models.FormField> AsDictionary() { throw null; }
         public float AsFloat() { throw null; }
+        public Azure.AI.FormRecognizer.Models.FormSelectionMarkState AsFormSelectionMarkState() { throw null; }
         public long AsInt64() { throw null; }
         public System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormField> AsList() { throw null; }
         public string AsPhoneNumber() { throw null; }
@@ -143,6 +144,7 @@ internal FormPage() { }
         public float Height { get { throw null; } }
         public System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormLine> Lines { get { throw null; } }
         public int PageNumber { get { throw null; } }
+        public System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormSelectionMark> SelectionMarks { get { throw null; } }
         public System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormTable> Tables { get { throw null; } }
         public float TextAngle { get { throw null; } }
         public Azure.AI.FormRecognizer.Models.LengthUnit Unit { get { throw null; } }
@@ -180,21 +182,36 @@ public static partial class FormRecognizerModelFactory
         public static Azure.AI.FormRecognizer.Models.FieldValue FieldValueWithInt64ValueType(long value) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FieldValue FieldValueWithListValueType(System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormField> value) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FieldValue FieldValueWithPhoneNumberValueType(string value) { throw null; }
+        public static Azure.AI.FormRecognizer.Models.FieldValue FieldValueWithSelectionMarkValueType(Azure.AI.FormRecognizer.Models.FormSelectionMarkState value) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FieldValue FieldValueWithStringValueType(string value) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FieldValue FieldValueWithTimeValueType(System.TimeSpan value) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormField FormField(string name, Azure.AI.FormRecognizer.Models.FieldData labelData, Azure.AI.FormRecognizer.Models.FieldData valueData, Azure.AI.FormRecognizer.Models.FieldValue value, float confidence) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormLine FormLine(Azure.AI.FormRecognizer.Models.FieldBoundingBox boundingBox, int pageNumber, string text, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormWord> words) { throw null; }
+        [System.ComponentModel.EditorBrowsableAttribute(System.ComponentModel.EditorBrowsableState.Never)]
         public static Azure.AI.FormRecognizer.Models.FormPage FormPage(int pageNumber, float width, float height, float textAngle, Azure.AI.FormRecognizer.Models.LengthUnit unit, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormLine> lines, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormTable> tables) { throw null; }
+        public static Azure.AI.FormRecognizer.Models.FormPage FormPage(int pageNumber, float width, float height, float textAngle, Azure.AI.FormRecognizer.Models.LengthUnit unit, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormLine> lines, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormTable> tables, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormSelectionMark> selectionMarks) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormPageCollection FormPageCollection(System.Collections.Generic.IList<Azure.AI.FormRecognizer.Models.FormPage> list) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormPageRange FormPageRange(int firstPageNumber, int lastPageNumber) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormRecognizerError FormRecognizerError(string errorCode, string message) { throw null; }
+        public static Azure.AI.FormRecognizer.Models.FormSelectionMark FormSelectionMark(Azure.AI.FormRecognizer.Models.FieldBoundingBox boundingBox, int pageNumber, string text, float confidence, Azure.AI.FormRecognizer.Models.FormSelectionMarkState state) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormTable FormTable(int pageNumber, int columnCount, int rowCount, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormTableCell> cells) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormTableCell FormTableCell(Azure.AI.FormRecognizer.Models.FieldBoundingBox boundingBox, int pageNumber, string text, int columnIndex, int rowIndex, int columnSpan, int rowSpan, bool isHeader, bool isFooter, float confidence, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormElement> fieldElements) { throw null; }
         public static Azure.AI.FormRecognizer.Models.FormWord FormWord(Azure.AI.FormRecognizer.Models.FieldBoundingBox boundingBox, int pageNumber, string text, float confidence) { throw null; }
         public static Azure.AI.FormRecognizer.Models.RecognizedForm RecognizedForm(string formType, Azure.AI.FormRecognizer.Models.FormPageRange pageRange, System.Collections.Generic.IReadOnlyDictionary<string, Azure.AI.FormRecognizer.Models.FormField> fields, System.Collections.Generic.IReadOnlyList<Azure.AI.FormRecognizer.Models.FormPage> pages) { throw null; }
         public static Azure.AI.FormRecognizer.Models.RecognizedFormCollection RecognizedFormCollection(System.Collections.Generic.IList<Azure.AI.FormRecognizer.Models.RecognizedForm> list) { throw null; }
         public static Azure.AI.FormRecognizer.Training.TrainingDocumentInfo TrainingDocumentInfo(string name, int pageCount, System.Collections.Generic.IEnumerable<Azure.AI.FormRecognizer.Models.FormRecognizerError> errors, Azure.AI.FormRecognizer.Training.TrainingStatus status) { throw null; }
     }
+    public partial class FormSelectionMark : Azure.AI.FormRecognizer.Models.FormElement
+    {
+        internal FormSelectionMark() { }
+        public float Confidence { get { throw null; } }
+        public Azure.AI.FormRecognizer.Models.FormSelectionMarkState State { get { throw null; } }
+    }
+    public enum FormSelectionMarkState
+    {
+        Selected = 0,
+        Unselected = 1,
+    }
     public partial class FormTable
     {
         internal FormTable() { }

sdk/formrecognizer/Azure.AI.FormRecognizer/samples/README.md

@@ -13,7 +13,7 @@ description: Samples for the Azure.AI.FormRecognizer client library
 # Azure Form Recognizer client SDK Samples
 Azure Cognitive Services Form Recognizer is a cloud service that uses machine learning to recognize form fields, text, and tables in form documents. It includes the following capabilities:
 
-- Recognize form content - Recognize and extract tables, lines, and words in forms documents, without the need to train a model.
+- Recognize form content - Recognize and extract tables, lines, words, and selection marks like radio buttons and check boxes in forms documents, without the need to train a model.
 - Recognize custom forms - Recognize and extract form fields and other content from your custom forms, using models you trained with your own form types.
 - Recognize receipts - Recognize and extract common fields from US receipts, using a pre-trained receipt model.
 

sdk/formrecognizer/Azure.AI.FormRecognizer/samples/Sample1_RecognizeFormContent.md

@@ -1,6 +1,6 @@
 # Recognize form content
 
-This sample demonstrates how to recognize tables, lines, and words in documents, without the need to train a model.
+This sample demonstrates how to recognize tables, lines, words, and selection marks like radio buttons and check boxes in forms documents, without the need to train a model.
 
 To get started you'll need a Cognitive Services resource or a Form Recognizer resource.  See [README][README] for prerequisites and instructions.
 
@@ -42,6 +42,17 @@ foreach (FormPage page in formPages)
             Console.WriteLine($"    Cell ({cell.RowIndex}, {cell.ColumnIndex}) contains text: '{cell.Text}'.");
         }
     }
+
+    for (int i = 0; i < page.SelectionMarks.Count; i++)
+    {
+        FormSelectionMark selectionMark = page.SelectionMarks[i];
+        Console.WriteLine($"Selection Mark {i} is {selectionMark.State.ToString()}.");
+        Console.WriteLine("        Its bounding box is:");
+        Console.WriteLine($"        Upper left => X: {selectionMark.BoundingBox[0].X}, Y= {selectionMark.BoundingBox[0].Y}");
+        Console.WriteLine($"        Upper right => X: {selectionMark.BoundingBox[1].X}, Y= {selectionMark.BoundingBox[1].Y}");
+        Console.WriteLine($"        Lower right => X: {selectionMark.BoundingBox[2].X}, Y= {selectionMark.BoundingBox[2].Y}");
+        Console.WriteLine($"        Lower left => X: {selectionMark.BoundingBox[3].X}, Y= {selectionMark.BoundingBox[3].Y}");
+    }
 }
 ```
 

sdk/formrecognizer/Azure.AI.FormRecognizer/src/FieldValue.cs

@@ -101,6 +101,17 @@ internal FieldValue(IReadOnlyDictionary<string, FormField> value)
             ValueDictionary = value;
         }
 
+        /// <summary>
+        /// Initializes a new instance of the <see cref="FieldValue"/> structure.
+        /// </summary>
+        /// <param name="value">The actual field value.</param>
+        internal FieldValue(FormSelectionMarkState value)
+            : this()
+        {
+            ValueType = FieldValueType.SelectionMark;
+            ValueSelectionMark = value;
+        }
+
         /// <summary>
         /// The data type of the field value.
         /// </summary>
@@ -150,6 +161,12 @@ internal FieldValue(IReadOnlyDictionary<string, FormField> value)
         /// </summary>
         private IReadOnlyDictionary<string, FormField> ValueDictionary { get; }
 
+        /// <summary>
+        /// The <see cref="FieldValueSelectionMark"/> value of this instance. Values are usually extracted from
+        /// <see cref="_fieldValue"/>, so this property is exclusively used for mocking.
+        /// </summary>
+        private FormSelectionMarkState ValueSelectionMark { get; }
+
         /// <summary>
         /// Gets the value of the field as a <see cref="string"/>.
         /// </summary>
@@ -349,5 +366,30 @@ public IReadOnlyDictionary<string, FormField> AsDictionary()
 
             return fieldDictionary;
         }
+
+        /// <summary>
+        /// Gets the value of the field as a <see cref="FormSelectionMarkState"/>.
+        /// </summary>
+        /// <returns>The value of the field converted to <see cref="FormSelectionMarkState"/>.</returns>
+        /// <exception cref="InvalidOperationException">Thrown when <see cref="ValueType"/> is not <see cref="FieldValueType.SelectionMark"/>.</exception>
+        public FormSelectionMarkState AsFormSelectionMarkState()
+        {
+            if (ValueType != FieldValueType.SelectionMark)
+            {
+                throw new InvalidOperationException($"Cannot get field as SelectionMark.  Field value's type is {ValueType}.");
+            }
+
+            if (_fieldValue == null)
+            {
+                return ValueSelectionMark;
+            }
+
+            if (!_fieldValue.ValueSelectionMark.HasValue)
+            {
+                throw new InvalidOperationException($"Field value is null.");
+            }
+
+            return _fieldValue.ValueSelectionMark.Value;
+        }
     }
 }

sdk/formrecognizer/Azure.AI.FormRecognizer/src/FieldValue_internal.cs

@@ -1,6 +1,8 @@
 // Copyright (c) Microsoft Corporation. All rights reserved.
 // Licensed under the MIT License.
 
+using System;
+using System.Collections.Generic;
 using Azure.Core;
 
 namespace Azure.AI.FormRecognizer.Models
@@ -11,11 +13,57 @@ internal partial class FieldValue_internal
         /// <summary>Integer value.</summary>
         public long? ValueInteger { get; }
 
+        /// <summary> Selection mark value. </summary>
+        public FormSelectionMarkState? ValueSelectionMark { get; }
+
         internal FieldValue_internal(string value)
         {
             Type = FieldValueType.String;
             ValueString = value;
             Text = value;
         }
+
+        /// <summary> Initializes a new instance of FieldValue_internal. </summary>
+        /// <param name="type"> Type of field value. </param>
+        /// <param name="valueString"> String value. </param>
+        /// <param name="valueDate"> Date value. </param>
+        /// <param name="valueTime"> Time value. </param>
+        /// <param name="valuePhoneNumber"> Phone number value. </param>
+        /// <param name="valueNumber"> Floating point value. </param>
+        /// <param name="valueInteger"> Integer value. </param>
+        /// <param name="valueArray"> Array of field values. </param>
+        /// <param name="valueObject"> Dictionary of named field values. </param>
+        /// <param name="valueSelectionMark"> Selection mark value. </param>
+        /// <param name="text"> Text content of the extracted field. </param>
+        /// <param name="boundingBox"> Bounding box of the field value, if appropriate. </param>
+        /// <param name="confidence"> Confidence score. </param>
+        /// <param name="elements"> When includeTextDetails is set to true, a list of references to the text elements constituting this field. </param>
+        /// <param name="page"> The 1-based page number in the input document. </param>
+        internal FieldValue_internal(FieldValueType type, string valueString, DateTimeOffset? valueDate, TimeSpan? valueTime, string valuePhoneNumber, float? valueNumber, long? valueInteger, IReadOnlyList<FieldValue_internal> valueArray, IReadOnlyDictionary<string, FieldValue_internal> valueObject, FormSelectionMarkState? valueSelectionMark, string text, IReadOnlyList<float> boundingBox, float? confidence, IReadOnlyList<string> elements, int? page)
+        {
+            Type = type;
+            ValueString = valueString;
+            ValueDate = valueDate;
+            ValueTime = valueTime;
+            ValuePhoneNumber = valuePhoneNumber;
+            ValueNumber = valueNumber;
+            ValueInteger = valueInteger;
+            ValueArray = valueArray;
+            ValueObject = valueObject;
+            ValueSelectionMark = valueSelectionMark;
+
+            BoundingBox = boundingBox;
+            Confidence = confidence;
+            Elements = elements;
+            Page = page;
+
+            if (Type == FieldValueType.SelectionMark)
+            {
+                ValueSelectionMark = FormSelectionMarkStateExtensions.ToFormSelectionMarkState(text);
+                Text = ValueSelectionMark.ToString();
+            }
+            else
+                Text = text;
+        }
     }
 }