initiated

app.py

@@ -0,0 +1,85 @@
+import streamlit as st
+import cv2
+import numpy as np
+from PIL import Image, ImageDraw
+# import imutils
+# import easyocr
+# import os
+# import pathlib
+# import platform
+# from xyxy_converter import yolov5_to_image_coordinates
+# import shutil
+from extractor import get_card_xy, get_digit
+
+# system_platform = platform.system()
+# if system_platform == 'Windows': pathlib.PosixPath = pathlib.WindowsPath
+
+# CUR_DIR = os.getcwd()
+# YOLO_PATH = f"{CUR_DIR}/yolov5"
+# MODEL_PATH = "runs/train/exp/weights/best.pt"
+
+def main():
+    st.title("Card number extractor")
+
+    # Use st.camera to capture images from the user's camera
+    img_file_buffer = st.camera_input(label='Please, take a photo of a card')
+
+    # try:
+    #     image = Image.open(img_file_buffer)
+    # except:
+    #     st.write('No shot detected')
+
+    # Check if an image is captured
+    if img_file_buffer is not None:
+        # Convert the image to a NumPy array
+        image = Image.open(img_file_buffer)
+        image_np = np.array(image)
+        resized_image = cv2.resize(image_np, (128, 128))
+        resized_image = resized_image.astype(np.uint8)
+        resized_image = cv2.cvtColor(resized_image, cv2.COLOR_BGR2RGB)
+        cv2.imwrite('card_image.jpg', resized_image)
+
+        # original_img = cv2.imread('card_image.jpg')
+        gray = cv2.cvtColor(resized_image, cv2.COLOR_BGR2GRAY)
+
+        x1, y1, x2, y2, card_confidence = get_card_xy(
+            model_path='credit_card_number_detector.tflite',
+            image_path='card_image.jpg'
+        )
+
+        st.write(card_confidence)
+
+        if card_confidence == 0:
+            display_text = "A card is not detected in the image!!!"
+            st.image('card_image.jpg', caption=f"{display_text}", use_column_width=True)
+        else:
+            cropped_image = gray[y1:y2, x1:x2]
+            # cropped_image = resized_image[y1:y2, x1:x2]
+            cropped_image = cv2.resize(cropped_image, (128, 128))
+            cv2.imwrite('card_number_image.jpg', cropped_image)
+            
+            extracted_digit = get_digit(
+                model_path="card_number_extractor.tflite", 
+                image_path='card_number_image.jpg', 
+                threshold=0.4
+            )
+
+            display_text = f'Here is the zoomed card number: {extracted_digit}'
+            st.image('card_number_image.jpg', caption=f"{display_text}", use_column_width=True)
+
+            image = Image.open('card_image.jpg')
+            image_resized = image.resize((640, 640))
+            draw = ImageDraw.Draw(image_resized)
+            draw.rectangle([x1, y1, x2, y2], outline="red", width=2)
+            class_name = 'card'
+            text = f"Class: {class_name}, Confidence: {card_confidence:.2f}"
+            draw.text((x1, y1), text, fill="red")
+            # Saving Images
+            image_resized.save('card_highlighted_image.jpg')
+            display_text = 'Here is the card on the image.'
+            st.image('card_highlighted_image.jpg', caption=f"{display_text}", use_column_width=True)
+
+        st.session_state.pop("card")
+
+if __name__ == "__main__":
+    main()

card_number_extractor.tflite

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:040573544cc95efd2efda024d4d5d526e18b1ec31afd894be877b1ca91627d37
+size 12102383

credit_card_number_detector.tflite

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce560776ea2d978fb398e2e80add8f8ff25dafecd299c989e873345971024aa0
+size 12104152

extractor.py

@@ -0,0 +1,159 @@
+import tensorflow as tf
+import numpy as np
+from PIL import Image
+import cv2
+
+def get_card_xy(model_path, image_path):
+    #model_path = 'odo_detector.tflite'
+    interpreter = tf.lite.Interpreter(model_path=model_path)
+    interpreter.allocate_tensors()
+
+    input_details = interpreter.get_input_details()
+    output_details = interpreter.get_output_details()
+
+    # Obtain the height and width of the corresponding image from the input tensor
+    image_height = input_details[0]['shape'][1] # 640
+    image_width = input_details[0]['shape'][2] # 640
+
+    # Image Preparation
+    # image_name = 'car.jpg'
+    image = Image.open(image_path)
+    image_resized = image.resize((image_width, image_height)) # Resize the image to the corresponding size of the input tensor and store it in a new variable
+
+    image_np = np.array(image_resized) #
+    image_np = np.true_divide(image_np, 255, dtype=np.float32) 
+    image_np = image_np[np.newaxis, :]
+
+    # inference
+    interpreter.set_tensor(input_details[0]['index'], image_np)
+    interpreter.invoke()
+
+    # Obtaining output results
+    output = interpreter.get_tensor(output_details[0]['index'])
+    output = output[0]
+    output = output.T
+
+    boxes_xywh = output[:, :4] #Get coordinates of bounding box, first 4 columns of output tensor
+    scores = output[:, 4]#np.max(output[..., 5:], axis=1) #Get score value, 5th column of output tensor
+    classes = np.zeros(len(scores))#np.argmax(output[..., 5:], axis=1) # Get the class value, get the 6th and subsequent columns of the output tensor, and store the largest value in the output tensor.
+
+    # Threshold Setting
+    # threshold = 0.7
+    final_score = 0
+    x_center, y_center, width, height = 0, 0, 0, 0
+    class_name = 'card_number'
+
+    # Bounding boxes, scores, and classes are drawn on the image
+    # draw = ImageDraw.Draw(image_resized)
+
+    for box, score, cls in zip(boxes_xywh, scores, classes):
+        if score >= final_score:
+            x_center, y_center, width, height = box
+            final_score = score
+            class_name = cls
+        else:
+            pass
+            
+    x1 = int((x_center - width / 2) * image_width)
+    y1 = int((y_center - height / 2) * image_height)
+    x2 = int((x_center + width / 2) * image_width)
+    y2 = int((y_center + height / 2) * image_height)
+
+    # draw.rectangle([x1, y1, x2, y2], outline="red", width=2)
+    # text = f"Class: {class_name}, Score: {final_score:.2f}"
+    # draw.text((x1, y1), text, fill="red")
+
+     # Saving Images
+    # image_resized.save('test_img.jpg')
+
+    return x1, y1, x2, y2, final_score
+
+def get_digit(model_path, image_path, threshold=0.5):
+
+    interpreter = tf.lite.Interpreter(model_path=model_path)
+    interpreter.allocate_tensors()
+
+    input_details = interpreter.get_input_details()
+    output_details = interpreter.get_output_details()
+
+    # Obtain the height and width of the corresponding image from the input tensor
+    image_height = input_details[0]['shape'][1] # 640
+    image_width = input_details[0]['shape'][2] # 640
+
+    # Image Preparation
+    # image_name = 'car.jpg'
+    # image = Image.open(image_path2)
+    # image_resized = image.resize((image_width, image_height)) # Resize the image to the corresponding size of the input tensor and store it in a new variable
+    image = cv2.imread(image_path)
+    # image_resized = np.resize(image, (image_width, image_height, 3))
+
+    image_np = np.array(image) #
+    image_np = np.true_divide(image_np, 255, dtype=np.float32) 
+    image_np = image_np[np.newaxis, :]
+
+    # inference
+    interpreter.set_tensor(input_details[0]['index'], image_np)
+    interpreter.invoke()
+
+    # Obtaining output results
+    output = interpreter.get_tensor(output_details[0]['index'])
+    output = output[0]
+    output = output.T
+
+    boxes_xywh = output[:, :4] #Get coordinates of bounding box, first 4 columns of output tensor
+    scores = np.max(output[:, 4:], axis=1) #Get score value, 5th column of output tensor
+    classes = np.argmax(output[:, 4:], axis=1) # Get the class value, get the 6th and subsequent columns of the output tensor, and store the largest value in the output tensor.
+
+    pred_list = []
+
+    prob_threshold = threshold
+
+    for box, score, cls in zip(boxes_xywh, scores, classes):
+
+        if score < prob_threshold:
+            continue
+
+        x_center, y_center, width, height = box
+        x1 = int((x_center - width / 2) * image_width)
+        y1 = int((y_center - height / 2) * image_height)
+        x2 = int((x_center + width / 2) * image_width)
+        y2 = int((y_center + height / 2) * image_height)
+
+        pred_list.append((x1, x2, cls, score))
+
+    pred_list = sorted(pred_list, key=lambda x: x[0])
+
+    num_list = []
+
+    temp_pred_list =[]
+
+    x_prev = 0
+
+    x_diff = min([elem[1] - elem[0] for elem in pred_list]) - 10
+
+    for idx, pred in enumerate(pred_list):
+    
+        if idx == 0:
+            temp_pred_list.append(pred)
+            x_prev = pred[0]
+        elif idx == len(pred_list) - 1:
+            temp_final_num = sorted(temp_pred_list, key=lambda x: x[-1], reverse=True)[0]
+            num_list.append(temp_final_num)
+        elif pred[0] - x_prev < x_diff:
+            temp_pred_list.append(pred)
+            x_prev = pred[0]
+        else:
+            temp_final_num = sorted(temp_pred_list, key=lambda x: x[-1], reverse=True)[0]
+            num_list.append(temp_final_num)
+            temp_pred_list = []
+            x_prev = pred[0]
+            temp_pred_list.append(pred)
+
+    sorted_number_list = sorted(num_list, key=lambda x: x[0])
+    # sorted_number_list = sorted(sorted_number_list, reverse=True, key= lambda x: x[-1])
+    # output_digit = float(''.join([str(int(i[2])) if i[2]!=10 else '.' for i in sorted_number_list]))
+    output_digit = float(''.join([str(int(i[2])) if i[2]!=10 else '.' for i in sorted_number_list]))
+    # output_digit = ''.join([str(int(i[2])) if i[2]!=10 else '.' for i in sorted_number_list[:10]])
+
+    return output_digit
+   

requirements.txt.txt

@@ -0,0 +1,70 @@
+absl-py==2.1.0
+altair==5.2.0
+astunparse==1.6.3
+attrs==23.2.0
+blinker==1.7.0
+cachetools==5.3.2
+certifi==2023.11.17
+charset-normalizer==3.3.2
+click==8.1.7
+colorama==0.4.6
+flatbuffers==23.5.26
+gast==0.5.4
+gitdb==4.0.11
+GitPython==3.1.41
+google-auth==2.26.2
+google-auth-oauthlib==1.2.0
+google-pasta==0.2.0
+grpcio==1.60.0
+h5py==3.10.0
+idna==3.6
+importlib-metadata==7.0.1
+Jinja2==3.1.3
+jsonschema==4.21.1
+jsonschema-specifications==2023.12.1
+keras==2.15.0
+libclang==16.0.6
+Markdown==3.5.2
+markdown-it-py==3.0.0
+MarkupSafe==2.1.4
+mdurl==0.1.2
+ml-dtypes==0.2.0
+numpy==1.26.3
+oauthlib==3.2.2
+opencv-python==4.9.0.80
+opt-einsum==3.3.0
+packaging==23.2
+pandas==2.2.0
+pillow==10.2.0
+protobuf==4.23.4
+pyarrow==15.0.0
+pyasn1==0.5.1
+pyasn1-modules==0.3.0
+pydeck==0.8.1b0
+Pygments==2.17.2
+python-dateutil==2.8.2
+pytz==2023.3.post1
+referencing==0.32.1
+requests==2.31.0
+requests-oauthlib==1.3.1
+rich==13.7.0
+rpds-py==0.17.1
+rsa==4.9
+six==1.16.0
+smmap==5.0.1
+streamlit==1.30.0
+tenacity==8.2.3
+tensorflow==2.15.0
+termcolor==2.4.0
+toml==0.10.2
+toolz==0.12.0
+tornado==6.4
+typing_extensions==4.9.0
+tzdata==2023.4
+tzlocal==5.2
+urllib3==2.1.0
+validators==0.22.0
+watchdog==3.0.0
+Werkzeug==3.0.1
+wrapt==1.14.1
+zipp==3.17.0