Add a SEMI OCR example

Author: yushulx

examples/official/semi_font_ocr/README.md

@@ -0,0 +1,32 @@
+# Semi-OCR Example with Dynamsoft Capture Vision SDK
+
+This example demonstrates how to use Dynamsoft Capture Vision SDK to recognize **SEMI (Semiconductor Equipment and Materials International)** font markings on silicon wafers. The custom model provided by Dynamsoft is trained for **single-density dot matrix fonts** and supports only **26 uppercase letters and 10 digits**.
+
+## Prerequisites
+
+- Python 3.8 or later
+- Install required packages:
+
+    ```bash
+    pip install -r requirements.txt
+    ```
+
+- Obtain a [30-day trial license](https://www.dynamsoft.com/customer/license/trialLicense/?product=dcv&package=cross-platform)
+
+## Usage
+
+1. Set your license key in `read_semi_ocr.py`:
+
+    ```python
+    err_code, err_str = LicenseManager.init_license("LICENSE-KEY")
+    ```
+
+2. Run the script:
+
+    ```bash
+    python read_semi_ocr.py
+    ```
+
+## Support
+
+For model-related issues, please contact [Dynamsoft Support](https://www.dynamsoft.com/company/customer-service/).

examples/official/semi_font_ocr/images/1LOZ0TTXM6.png

@@ -0,0 +1,52 @@
+import os
+from dynamsoft_capture_vision_bundle import *
+
+if __name__ == '__main__':
+    # 1.Initializes license.
+    # You can request or extend a trial license at: https://www.dynamsoft.com/customer/license/trialLicense?product=dcv&utm_source=samples&package=python
+    # The string below is a free public trial license. Note: an active internet connection is required for this license to work.
+    err_code, err_str = LicenseManager.init_license("DLS2eyJoYW5kc2hha2VDb2RlIjoiMjAwMDAxLTE2NDk4Mjk3OTI2MzUiLCJvcmdhbml6YXRpb25JRCI6IjIwMDAwMSIsInNlc3Npb25QYXNzd29yZCI6IndTcGR6Vm05WDJrcEQ5YUoifQ==")
+    if err_code != EnumErrorCode.EC_OK and err_code != EnumErrorCode.EC_LICENSE_CACHE_USED:
+        print("License initialization failed: ErrorCode: " + str(err_code) + ", ErrorString: " + err_str)
+    else:
+        # 2.Creates an instance of CaptureVisionRouter.
+        cvr = CaptureVisionRouter()
+        with open('models/semi-ocr.data', 'rb') as f:
+            model_data = f.read()
+            
+        err_code, err_str = cvr.append_model_buffer('semi-ocr', model_data, 1)
+        print("Model initialization: ErrorCode: " + str(err_code) + ", ErrorString: " + err_str)
+        
+        # 3.Performs capture jobs(recognizing text lines) on an image
+        err_code, err_str = cvr.init_settings_from_file("semi-ocr.json")
+        print("Template initialization: ErrorCode: " + str(err_code) + ", ErrorString: " + err_str)
+        
+        files_list = []
+        while True:
+            path = input("Please enter the image file/directory path: ").strip('\'"')
+            if not os.path.exists(path):
+                print("File not found: " + path)
+            else:
+                if os.path.isfile(path):
+                    files_list.append(path)
+                elif os.path.isdir(path):
+                    files =os.listdir(path)
+                    for file in files:
+                        if file.endswith(".jpg") or file.endswith(".jpeg") or file.endswith(".png"):
+                            files_list.append(os.path.join(path, file))
+                break
+            
+        for file_path in files_list:
+            result=cvr.capture(file_path, "recognize_semi_ocr")
+
+            print("File: " + file_path)
+            
+            # 4.Outputs the result.
+            if result.get_error_code() != EnumErrorCode.EC_OK:
+                print("Error: " + str(result.get_error_code())+ result.get_error_string())
+            else:
+                items = result.get_items()
+                # print(f'read {len(items)} items')
+                for item in items:
+                    if isinstance(item, TextLineResultItem):
+                        print(item.get_text())

examples/official/semi_font_ocr/images/5EVZFWOM0E.png

@@ -0,0 +1,2 @@
+dynamsoft-capture-vision-bundle
+opencv-python

examples/official/semi_font_ocr/images/C2Q1T26TUK.png

@@ -0,0 +1,147 @@
+{
+  "CaptureVisionTemplates": [
+    {
+      "Name": "recognize_semi_ocr",
+      "ImageROIProcessingNameArray": [
+        "roi_test"
+      ],
+      "Timeout": 2000
+    }
+  ],
+  "TargetROIDefOptions": [
+    {
+      "Name": "roi_test",
+      "TaskSettingNameArray": [
+        "task_test"
+      ]
+    }
+  ],
+  "TextLineSpecificationOptions": [
+    {
+      "Name": "tls_test",
+      "CharacterModelName": "",
+      "TextLineRecModelName": "semi-ocr",
+      "CharHeightRange": [
+        1,
+        1000,
+        1
+      ],
+      "StringLengthRange": [
+        7,
+        13
+      ],
+      "BinarizationModes": [
+        {
+          "BlockSizeX": 30,
+          "BlockSizeY": 30,
+          "Mode": "BM_LOCAL_BLOCK",
+          "EnableFillBinaryVacancy": 0,
+          "ThresholdCompensation": 15
+        }
+      ]
+    }
+  ],
+  "LabelRecognizerTaskSettingOptions": [
+    {
+      "Name": "task_test",
+      "TextLineSpecificationNameArray": [
+        "tls_test"
+      ],
+      "SectionArray": [
+        {
+          "Section": "ST_REGION_PREDETECTION",
+          "ImageParameterName": "ip_test",
+          "StageArray": [
+            {
+              "Stage": "SST_PREDETECT_REGIONS"
+            }
+          ]
+        },
+        {
+          "Section": "ST_TEXT_LINE_LOCALIZATION",
+          "ImageParameterName": "ip_test",
+          "StageArray": [
+            {
+              "Stage": "SST_LOCALIZE_TEXT_LINES"
+            }
+          ]
+        },
+        {
+          "Section": "ST_TEXT_LINE_RECOGNITION",
+          "ImageParameterName": "ip_test",
+          "StageArray": [
+            {
+              "Stage": "SST_RECOGNIZE_RAW_TEXT_LINES"
+            },
+            {
+              "Stage": "SST_ASSEMBLE_TEXT_LINES"
+            }
+          ]
+        }
+      ]
+    }
+  ],
+  "CaptureVisionModelOptions": [
+    {
+      "Name": "semi-ocr",
+      "MaxModelInstances": 1
+    }
+  ],
+  "ImageParameterOptions": [
+    {
+      "Name": "ip_test",
+      "ApplicableStages": [
+        {
+          "Stage": "SST_TRANSFORM_GRAYSCALE",
+          "GrayscaleTransformationModes": [
+            {
+              "Mode": "GTM_ORIGINAL"
+            },
+            {
+              "Mode": "GTM_INVERTED"
+            }
+          ]
+        },
+        {
+          "Stage": "SST_DETECT_TEXTURE",
+          "TextureDetectionModes": [
+            {
+              "Mode": "TDM_GENERAL_WIDTH_CONCENTRATION",
+              "Sensitivity": 8
+            }
+          ]
+        },
+        {
+          "Stage": "SST_BINARIZE_IMAGE",
+          "BinarizationModes": [
+            {
+              "EnableFillBinaryVacancy": 0,
+              "ThresholdCompensation": 21,
+              "Mode": "BM_LOCAL_BLOCK",
+              "MorphOperation": "Dilate",
+              "MorphOperationKernelSizeX": 7,
+              "MorphOperationKernelSizeY": 7
+            }
+          ]
+        },
+        {
+          "Stage": "SST_DETECT_TEXT_ZONES",
+          "TextDetectionMode": {
+            "Mode": "TTDM_LINE",
+            "CharHeightRange": [
+              5,
+              1000,
+              1
+            ],
+            "StringLengthRange": [
+              7,
+              13
+            ],
+            "Direction": "HORIZONTAL",
+            "Sensitivity": 7
+          }
+        }
+      ]
+    }
+  ]
+}