updated model and added logs

conf/application.conf

@@ -11,4 +11,4 @@ host=localhost
 port=8200
 
 [PATH]
-model_path = scripts/utility/model/yolov8l_r1_r2_lp_best.pt
+model_path = scripts/utility/model/r1_r2_lpr_yolov8s.pt

requirements.txt

@@ -55,4 +55,5 @@ uvicorn==0.19.0
 ultralytics
 paddleocr==2.7.0.3
 paddlepaddle==2.5.2
+#loguru==0.7.2
 

scripts/core/handler/Text_Extraction_Handler.py

@@ -3,103 +3,113 @@ from ultralytics import YOLO
 import torch
 import cv2
 from scripts.config.app_configurations import *
+from scripts.logging.application_logging import logger
 
 device = 'cuda' if torch.cuda.is_available() else 'cpu'
-# print(f'Using device: {device}')
+
 def load_ocr_model():
+
+    """
+    Load paddle OCR model
+
+    """
     ocr = PaddleOCR(use_angle_cls=True, lang='en')
+    logger.info("Loaded OCR model")
     return ocr
 
 def load_yolo_model():
+    """
+    Load license plate detection model
+
+    """
     model = YOLO(MODEL_PATH).to(device)
+    logger.info("Loaded Yolo model")
     return model
 
 
 def predict(img_path, ocr, yolo_model):
-    img = cv2.imread(img_path)
-    results = yolo_model.predict(img, iou=0.7)
-    if results:
-        r1_txt = None
-        r2_txt = None
-
-        for r in results:
-
-            # annotator = Annotator(img)
-
-            boxes = r.boxes
-            for box in boxes:
-                b = box.xyxy[0]  # get box coordinates in (top, left, bottom, right) format
-                # print("bb:  ",b.tolist())
-                x1 = int(b[0])
-                y1 = int(b[1])
-                x2 = int(b[2])
-                y2 = int(b[3])
-                w = abs(x2 - x1)
-                h = abs(y2 - y1)
-
-                cv2.rectangle(img, (x1, y1), (x2, y2), (255, 0, 0), 1)
-
-                # lp = img[max(y1 - 30, 0):min(y1+h + 30, img.shape[0]), max(x1 - 30, 0):min(x1+w + 30, img.shape[1])]
-
-                lp  = img[y1:y1 + h, x1:x1 + w]
-
-                c = box.cls
-
-                # print("class:   ", c)
-                # annotator.box_label(b, model.names[int(c)])
-
-                # img = annotator.result()
-                # cv2.imshow('YOLO V8 Detection', cv2.resize(img, (900, 600)))
-                if int(c) == 0:
-                    result = ocr.ocr(lp, cls=True)
-                    for idx in range(len(result)):
-                        res = result[idx]
-                        text_x = 10
-                        text_y = 30
-                        if res:
-
-                            for ind, line in enumerate(res):
-                                # print("line")
-                                # print(line)
-
-                                text_tuple = line[1]
-                                if r1_txt is None:
-                                    r1_txt = text_tuple[0]
-                                    print("R1  TEXT:   ", r1_txt)
-                                text_conf = text_tuple[1]
-                elif int(c) == 1:
-
-                    result = ocr.ocr(lp, cls=True)
-                    for idx in range(len(result)):
-                        res = result[idx]
-                        text_x = 10
-                        text_y = 30
-                        if res:
-
-                            for ind, line in enumerate(res):
-                                # print("line")
-                                # print(line)
-
-                                text_tuple = line[1]
-                                if r2_txt is None:
-                                    r2_txt = text_tuple[0]
-                                    print("R2   TEXT:   ", r2_txt)
-                                text_conf = text_tuple[1]
-                            # cv2.putText(lp, txt, (text_x , text_y + ind*50), cv2.FONT_HERSHEY_SIMPLEX , 1, (255, 0, 0), 2, cv2.LINE_AA)
-
-                # cv2.imshow('detected LPR', lp)
-                # cv2.waitKey(0)
-                #
-
-        if r1_txt is not None and r2_txt is not None:
-            LP = f"{r1_txt}  {r2_txt}"
-
-            # print("LP:  ", LP)
-            return LP
-
-    return "Image Not Clear"
-    # result = ocr.ocr(img_path, cls=False,det=True,rec=True)[0]
-    # txts = [line[1][0] for line in result]
+    try:
+
+        img = cv2.imread(img_path)
+
+        results = yolo_model.predict(img, iou=0.7) # predict r1, r2 and license plate using yolo model r1 - single letter region in license plate, r2 - four or five letter region in license plate
+
+        if results:
+            r1_txt = None
+            r2_txt = None
+
+            for r in results:
+
+
+
+                boxes = r.boxes
+                for box in boxes:
+                    b = box.xyxy[0]  # get box coordinates in (top, left, bottom, right) format
+                    # print("bb:  ",b.tolist())
+                    x1 = int(b[0])
+                    y1 = int(b[1])
+                    x2 = int(b[2])
+                    y2 = int(b[3])
+                    w = abs(x2 - x1)
+                    h = abs(y2 - y1)
+
+
+
+                    lp  = img[y1:y1 + h, x1:x1 + w]
+
+                    c = box.cls
+
+
+                    if int(c) == 0:
+                        #extracting text from r1
+                        # cv2.imwrite("LP.png", lp)
+                        result = ocr.ocr(lp, cls=True)
+                        for idx in range(len(result)):
+                            res = result[idx]
+
+                            if res:
+
+                                for ind, line in enumerate(res):
+
+
+                                    text_tuple = line[1]
+                                    if r1_txt is None:
+                                        r1_txt = text_tuple[0]
+                                        logger.info(f"r1 text : {r1_txt}")
+                                    text_conf = text_tuple[1]
+                    elif int(c) == 1:
+                        #extracting text from r2
+                        result = ocr.ocr(lp, cls=True)
+                        for idx in range(len(result)):
+                            res = result[idx]
+
+                            if res:
+
+                                for ind, line in enumerate(res):
+
+
+                                    text_tuple = line[1]
+                                    if r2_txt is None:
+                                        r2_txt = text_tuple[0]
+                                        logger.info(f"r2 text : {r2_txt}")
+                                    text_conf = text_tuple[1]
+                                # cv2.putText(lp, txt, (text_x , text_y + ind*50), cv2.FONT_HERSHEY_SIMPLEX , 1, (255, 0, 0), 2, cv2.LINE_AA)
+
+                    # cv2.imshow('detected LPR', lp)
+                    # cv2.waitKey(0)
+                    #
+
+            if r1_txt is not None and r2_txt is not None:
+                LP = f"{r1_txt}  {r2_txt}"
+
+                logger.info(f"License plate reading : {LP}")
+                return LP
+
+        return "Image Not Clear"
+    except Exception as e:
+        logger.info(f"Error while reading license plate : {e}")
+        return f"Error while reading license plate : {e}"
+
 
 
 

scripts/core/services/Text_Extraction_Service.py

@@ -8,6 +8,7 @@ from scripts.core.handler.Text_Extraction_Handler import load_ocr_model,load_yol
 from scripts.logging.application_logging import logger
 from scripts.config.app_constants import TEXT_EXTRACT
 
+
 ocr = load_ocr_model()
 yolo_model = load_yolo_model()
 router = APIRouter()