idvpackage 3.0.12__tar.gz → 3.0.14__tar.gz

{idvpackage-3.0.12/idvpackage.egg-info → idvpackage-3.0.14}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: idvpackage
-Version: 3.0.12
+Version: 3.0.14
 Summary: This repository contains a Python program designed to execute Optical Character Recognition (OCR) and Facial Recognition on images.
 Home-page: https://github.com/NymCard-Payments/project_idv_package
 Classifier: Programming Language :: Python :: 3

{idvpackage-3.0.12 → idvpackage-3.0.14}/idvpackage/common.py

@@ -284,10 +284,10 @@ def load_and_process_image_deepface(image_input, country=None):
                 if w < 40 or h < 50:
                     logging.info(f"Face too small for SDN: w={w}, h={h}")
                     return None, None, 0.0
-            else:
-                if w < 80 or h < 90:
-                    logging.info(f"Face too small: w={w}, h={h}")
-                    return None, None, 0.0
+            # else:
+            #     if w < 80 or h < 90:
+            #         logging.info(f"Face too small: w={w}, h={h}")
+            #         return None, None, 0.0
 
             # All checks passed
             return biggest_face, img_to_process, confidence
@@ -346,13 +346,14 @@ def load_and_process_image_deepface(image_input, country=None):
 
             if face_objs is None:
                 continue
-            
             else:
                 break  # Exit loop on first valid detection
 
             # Keep best fallback (just in case)
 
-        if best_face_objs is None or best_confidence < CONFIDENCE_THRESHOLD:
+
+        #remove best_confidence < CONFIDENCE_THRESHOLD to take best of all angles when confidence is less than threshold
+        if best_face_objs is None:
             print(f"No valid face found (threshold={CONFIDENCE_THRESHOLD})")
             return [], []
 

{idvpackage-3.0.12 → idvpackage-3.0.14}/idvpackage/iraq_id_extraction_withopenai.py

@@ -146,7 +146,7 @@ class IraqiPassport(BaseModel):
         ..., description="Date of birth as extracted from MRZ (in DD/MM/YYYY format)"
     )
     id_number_mrz: str = Field(
-        ..., description="ID number as extracted from MRZ"
+        ..., min_length=9, max_length=9, description="ID number as extracted from MRZ"
     )
     expiry_date_mrz: str = Field(
         ..., description="Expiry date as extracted from MRZ (in DD/MM/YYYY format)"

{idvpackage-3.0.12 → idvpackage-3.0.14}/idvpackage/ocr.py

@@ -896,34 +896,37 @@ class IdentityVerification:
             from idvpackage.ocr_utils import document_on_printed_paper
             from idvpackage.qatar_id_extraction import get_response_from_openai_qat
 
-            front_data = {"error": "", "doc_type": "national_identity_card"}
+            result = {"error": "", "doc_type": "national_identity_card"}
 
             try:
                 processed_front_id ,compressed_image_data= self.image_conversion_and_compression(front_id)
                 st = time.time()
-                front_data_fields = get_response_from_openai_qat(
+                
+                front_data = get_response_from_openai_qat(
                     compressed_image_data, "front", country, self.openai_key
                 )
                 logging.info(f"----------------Time taken for vision front: {time.time() - st} seconds\n")
-                logging.info(f"front_data_fields: {json.dumps(front_data_fields, indent=2, ensure_ascii=False)}")
+                logging.info(f"front_data: {json.dumps(front_data, indent=2, ensure_ascii=False)}")
 
-                front_data_fields['issuing_country'] = 'QAT'
+                front_data['issuing_country'] = 'QAT'
+
+                if not front_data.get("header_verified", False):
+                    result["error"] = "not_front_id"
+                    return result
 
-                if not front_data_fields["header_verified"]:
-                    front_data["error"] = "not_front_id"
-                    return front_data
-                
                 logging.info(  f"----------------Time taken for qatar OpenAI and  final extraction front: {time.time() - st} seconds\n")
 
-                expiry_date = front_data_fields.get("expiry_date", "")
-                if expiry_date:
-                    try:
-                        dt_obj = datetime.strptime(expiry_date, "%d/%m/%Y")
-                        tomorrow = datetime.today() + timedelta(days=1)
-                        if dt_obj < tomorrow:
-                            front_data["error"] = "expired_id"
-                    except:
-                        pass
+                expiry_date = front_data.get("expiry_date", "")
+                try:
+                    if expiry_date:
+                        logging.info(f"Extracted Expiry Date for expiry verification: {expiry_date}")
+                        from idvpackage.ocr_utils import is_expired_id
+                        if is_expired_id(expiry_date):
+                            result['error'] = 'expired_id'
+                            logging.info(f"ID is expired with expiry date: {expiry_date}")
+                            return result
+                except Exception as e:
+                    logging.info(f"Error in expiry date calculation: {e}")
 
                 image = np.array(processed_front_id)
                 doc_on_pp_result = document_on_printed_paper(image)
@@ -935,9 +938,9 @@ class IdentityVerification:
 
                 
                 if front_face_encodings is None or len(front_face_encodings) == 0:
-                    front_data['error'] = 'face_not_detected'
+                    result['error'] = 'face_not_detected'
                     logging.info("No face detected in front image")
-                    return front_data
+                    return result
                 
                 logging.info(f"----------------Time taken for face extraction front: {time.time() - st} seconds\n")
 
@@ -962,8 +965,8 @@ class IdentityVerification:
                     "front_face_encodings": front_face_encodings_str,
                 }
 
-                front_data_fields.update(front_data_temp)
-                front_data.update(front_data_fields)
+                front_data.update(front_data_temp)
+                result.update(front_data)
                 
                 required_keys = ["expiry_date", "name", "id_number"]
                 empty_string_keys = [
@@ -973,18 +976,18 @@ class IdentityVerification:
                 ]
 
                 if empty_string_keys:
-                    front_data["error"] = "missing_key_fields"
+                    result["error"] = "missing_key_fields"
 
                 if front_data.get("error"):
-                    return front_data
+                    return result
 
 
             except Exception as e:
                 logging.info(f"exception in QAT front ID extraction: {e}")
-                front_data["error"] = "bad_image"
-                front_data["error_details"] = e
+                result["error"] = "bad_image"
+                result["error_details"] = e
 
-            return front_data
+            return result
 
         if country == "LBN":
             logging.info("----------------Working on LBN\n")
@@ -1032,9 +1035,10 @@ class IdentityVerification:
                 logging.info(f"----------------Time taken for OpenAI and  final extraction front: {time.time() - st} seconds\n")
                 logging.info(f"Extracted LBN front data fields: {json.dumps(front_data_fields, indent=2, ensure_ascii=False)}")
 
-                if front_data_fields.get('header_verified') is not True:
+                if not front_data_fields.get("header_verified", False):
                     front_data["error"] = "not_front_id"
                     return front_data   
+                
                 if 'id_number' in front_data_fields:
                     id_number = front_data_fields['id_number']
                     if id_number and len(id_number) < 12:
@@ -1198,7 +1202,7 @@ class IdentityVerification:
                 logging.info(f"Extracted SDN front data fields: {json.dumps(front_data_fields, indent=2, ensure_ascii=False)}")
                 logging.info(f"----------------Time taken for OpenAI and  final extraction front: {time.time() - st} seconds\n")
 
-                if not front_data_fields['header_verified']:
+                if not front_data_fields.get("header_verified", False):
                     front_data_fields['error'] = 'not_front_id'
                     logging.error(f"ID not verified in the document data: {front_data_fields['header_verified']}")
                     return front_data_fields
@@ -1787,7 +1791,7 @@ class IdentityVerification:
  
         if country == "QAT":
             from idvpackage.qatar_id_extraction import get_response_from_openai_qat
-            back_data = {"error": "", "doc_type": "national_identity_card"}
+            result = {"error": "", "doc_type": "national_identity_card"}
             try:
                 processed_back_id = self.image_conversion(back_id)
                 compressed_image = BytesIO()
@@ -1797,11 +1801,15 @@ class IdentityVerification:
                 compressed_image_data = compressed_image.getvalue()
                 
                 st = time.time()
-                back_extraction_result = get_response_from_openai_qat(compressed_image_data, "back", country, self.openai_key)
+                back_data = get_response_from_openai_qat(compressed_image_data, "back", country, self.openai_key)
                 logging.info(f"----------------Time taken for OpenAI and  final extraction back: {time.time() - st} seconds\n")
-                logging.info(f"back_extraction_result: {json.dumps(back_extraction_result, ensure_ascii=False, indent=2)}")
-               
-                back_data.update(back_extraction_result)
+                logging.info(f"back_extraction_result: {json.dumps(back_data, ensure_ascii=False, indent=2)}")
+                
+                if not back_data.get('back_header_verified', False):
+                    result["error"] = "not_back_id"
+                    return result
+
+                result.update(back_data)
                 back_data_update = {
                     "back_extracted_data": "",
                     "back_coloured": True,
@@ -1812,13 +1820,13 @@ class IdentityVerification:
                     "back_glare": "clear",
                 }
 
-                back_data.update(back_data_update)
+                result.update(back_data_update)
                 
             except Exception as e:
-                back_data["error"] = "bad_image"
-                back_data["error_details"] = e
+                result["error"] = "bad_image"
+                result["error_details"] = e
 
-            return back_data
+            return result
 
         if country == "LBN":
             from idvpackage.blur_detection import is_image_blur
@@ -2339,7 +2347,7 @@ class IdentityVerification:
                 )
                 logging.info(f"passport_details: {json.dumps(passport_details, ensure_ascii=False, indent=2)}")
 
-                if not passport_details['header_verified'] :
+                if not passport_details.get('header_verified', False):
                     passport_data["error"] = "not_passport"
                     return passport_data
                 
@@ -2497,7 +2505,7 @@ class IdentityVerification:
                 logging.info(f"Passport details extracted: {json.dumps(passport_details, ensure_ascii=False, indent=2)}")
                 logging.info(f"----------------Time taken for OpenAI and  final extraction passport: {time.time() - st} seconds\n")
 
-                if not passport_details['header_verified']:
+                if not passport_details.get('header_verified', False):
                     passport_data["error"] = "not_passport"
                     return passport_data
             
@@ -2637,7 +2645,7 @@ class IdentityVerification:
                 )
                 logging.info(f"passport_details: {json.dumps(passport_details, ensure_ascii=False, indent=2)}")
 
-                if not passport_details['header_verified']:
+                if not passport_details.get('header_verified', False):
                     passport_data["error"] = "not_passport"
                     return passport_data
                 
@@ -2777,10 +2785,7 @@ class IdentityVerification:
                 logging.info(f"----------------Passport details: {json.dumps(passport_details, indent=4, ensure_ascii=False)}\n")
                 logging.info(f"----------------Time taken for openai final extraction passport: {time.time() - st} seconds\n")
 
-                if (
-                    passport_details.get("header_verified", "")
-                    and passport_details["header_verified"] == "False"
-                ):
+                if not passport_details.get("header_verified", False):
                     passport_data["error"] = "not_passport"
                     return passport_data
 
@@ -2909,7 +2914,7 @@ class IdentityVerification:
                     f"----------------Time taken for OpenAI and  final extraction passport: {time.time() - st} seconds\n"
                 )
 
-                if passport_details.get("header_verified", "") == "False":
+                if not passport_details.get("header_verified", False):
                     passport_data["error"] = "not_passport"
                     return passport_data
 

{idvpackage-3.0.12 → idvpackage-3.0.14}/idvpackage/ocr_utils.py

@@ -560,15 +560,16 @@ def data_consistency_check(data, front_id_text, back_id_text, country, back_img)
     if country == 'IRQ':
         data_consistency['breakdown']['document_type']['result'] = 'clear'
 
-        front_id_number = data.get('id_number', '')
-        back_id_number = data.get('id_number_mrz', '')
-        
-        logging.info(f'Front ID Number: {front_id_number}')
-        logging.info(f'Back ID Number: {back_id_number}')
+        if data.get('doc_type') == 'national_identity_card':
+            front_id_number = data.get('id_number', '')
+            back_id_number = data.get('id_number_mrz', '')
+            
+            logging.info(f'Front ID Number: {front_id_number}')
+            logging.info(f'Back ID Number: {back_id_number}')
 
-        if front_id_number and back_id_number:
-            if front_id_number != back_id_number:
-                data_consistency['breakdown']['multiple_data_sources_present']['result'] = 'consider'
+            if front_id_number and back_id_number:
+                if front_id_number != back_id_number:
+                    data_consistency['breakdown']['multiple_data_sources_present']['result'] = 'consider'
 
     if country == 'QAT':
         data_consistency['breakdown']['gender']['result'] = 'clear'
@@ -684,7 +685,7 @@ def get_name_match_mrz(data, doc_type):
         name = data.get("name", "")
 
     elif data['nationality'] == 'IRQ':
-        name = data.get("name", "")
+        name = data.get("full_name", "")
         logging.info(f"Name extracted for IRQ: {name}")
     
     elif data['nationality'] == 'PSE':

{idvpackage-3.0.12 → idvpackage-3.0.14}/idvpackage/qatar_id_extraction.py

@@ -23,7 +23,7 @@ Extract ALL fields from this Qatar National ID **front side** image with high ac
 
 Return a JSON object with the following fields (use the exact field names):
 
-    -id_number: The ID number exactly as shown on the card (preserve original format)
+    - id_number: The ID number exactly as shown on the card (preserve original format)
     - dob: Date of birth exactly as shown on the card, but always return in DD/MM/YYYY format (e.g., '15/06/1990'). If the card shows a different format, convert it to DD/MM/YYYY.
     - expiry_date: Date of expiry exactly as shown on the card, but always return in DD/MM/YYYY format (e.g., '15/06/1990'). If the card shows a different format, convert it to DD/MM/YYYY.
     - name: Full name in English as printed on the card (extract exactly as written)
@@ -49,8 +49,8 @@ Extract ALL fields from this Qatar National ID **back side** image with high acc
 Return a JSON object with the following fields (use the exact field names):
 
 - employer: Employer name in Arabic as printed on the card (extract exactly as written)
-- employer_en: Translate the Arabic employer name to English (e.g., شركة → Company, مؤسسة → Establishment)
-- passport_number: Passport number as printed on the card (extract exactly as written, e.g., EA0605652)
+- employer_en: Translate the Arabic employer name to English 
+- passport_number: Passport number as printed on the card (extract exactly as written)
 - passport_expiry: Passport expiry date  always in DD/MM/YYYY format.
 - back_header_verified: Return True if one of the texts present in the image is "Director General of the General Department" or "Directorate of Passports" or "Passport number" or "Serial"; otherwise False.
 - card_number: Serial number exactly as shown on the card (preserve original format)
@@ -64,7 +64,7 @@ Instructions:
 
 class QatarFront(BaseModel):
    
-    id_number: str = Field(...,min_length=9, max_length=11,
+    id_number: str = Field(...,min_length=11, max_length=11,
         description = "The ID number exactly as shown on the card (preserve original format)",
     )
     
@@ -114,29 +114,29 @@ class QatarFront(BaseModel):
     )
 class QatarBack(BaseModel):
     employer: str = Field(...,
-        description = "Employer name in Arabic (extract exactly as written on the card)",
+        description = "Employer name in Arabic (extract exactly as written on the card) return empty string if not present",
     )
 
     employer_en: str = Field(...,
-        description = "TRANSLATE the Arabic  employer name to English (e.g., شركة → Company, مؤسسة → Establishment)",
+        description = "TRANSLATE the Arabic  employer name to English, if employer is not present return empty string",
 
     )
 
     passport_number: str = Field(...,
-        description = "Passport number  extract exactly as written on the card ex: EA0605652."
+        description = "Passport number  extract exactly as written on the card, return empty string if not present",
     )
 
     passport_expiry: str = Field(...,
-        description = "Passport expiry date exactly as shown on the card (preserve (dd/mm/yyyy) format)",
+        description = "Passport expiry date exactly as shown on the card (preserve (dd/mm/yyyy) format), return empty string if not present",
     )
 
     back_header_verified: bool = Field(
         ...,
-        description=" Return True if one of the texts present in the image Director General of the General Department or Directorate of Passports or Passport number or Serial",
+        description=" Return True if one of the texts present in the image Director General of the General Department or Directorate of Passports or Passport number or Serial else return False",
     )
 
     card_number: str = Field(...,
-        description = "Serial number exactly as shown on the card (preserve original format)",
+        description = "Serial number exactly as shown on the card (preserve original format), return empty string if not present",
     )
 
 

{idvpackage-3.0.12 → idvpackage-3.0.14/idvpackage.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: idvpackage
-Version: 3.0.12
+Version: 3.0.14
 Summary: This repository contains a Python program designed to execute Optical Character Recognition (OCR) and Facial Recognition on images.
 Home-page: https://github.com/NymCard-Payments/project_idv_package
 Classifier: Programming Language :: Python :: 3

{idvpackage-3.0.12 → idvpackage-3.0.14}/setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = idvpackage
-version = 3.0.12
+version = 3.0.14
 description = This repository contains a Python program designed to execute Optical Character Recognition (OCR) and Facial Recognition on images.
 long_description = file: README.md
 long_description_content_type = text/markdown

{idvpackage-3.0.12 → idvpackage-3.0.14}/setup.py

@@ -9,7 +9,7 @@ with open("README.md", "r", encoding="utf-8") as fh:
 
 setup(
     name="idvpackage",
-    version="3.0.12",
+    version="3.0.14",
     description=(
         "This repository contains a Python program designed to execute Optical Character Recognition (OCR)"
         " and Facial Recognition on images."