app.py

# %%
# pip install torch torchvision torchaudio yolov5
import gradio as gr
import cv2
import requests
import os
import torch 
import numpy as np
from ultralytics import YOLO
import yolov5
import easyocr
from paddleocr import PaddleOCR
import keras_ocr
import matplotlib.pyplot as plt
from paddleocr import draw_ocr
import pandas as pd

# %%
# Loading Yolo V5 model
model = yolov5.load('License_Plate_Model_Y5.pt')#, device="cpu")

# %%
def get_LicencePlate(frame, conf_threshold: gr.inputs.Slider = 0.1):
    
    lst_plate_xyxy = []
    # Setting model configuration 
    model.conf = conf_threshold

    results = model(frame)
    for index, row in results.pandas().xyxy[0].iterrows():
        xmin = int(row['xmin'])
        ymin = int(row['ymin'])
        xmax = int(row['xmax'])
        ymax = int(row['ymax'])
        class_name=(row['name'])
        if (class_name == 'license-plate'):
            lst_plate_xyxy.append((xmin,ymin,xmax,ymax, class_name))

    return lst_plate_xyxy

# %%
def extract_License_Number_PaddleOCR(image_path, conf_threshold):
    image = cv2.imread(image_path)
    # image = cv2.resize(image,(1020,800))
    results = get_LicencePlate(image, conf_threshold)
    PaddleOCR_output = []
    # words = []
    # boxes = []
    # scores = []
    for index in results:
        x1 = index[0]
        y1 = index[1]
        x2 = index[2]
        y2 = index[3]
        class_name=(index[4])

        cropped_img = image[y1:y2, x1:x2]

        # processed_img = cropped_img
        processed_img =  cropped_img

        # Perform OCR using PaddleOCR
        paddle_ocr = PaddleOCR()
        paddle_ocr_result = paddle_ocr.ocr(processed_img)

        PaddleOCR_output.append(paddle_ocr_result)    

        print('\nPaddleOCR:')
        print(''.join([text[1][0] + ' ' for text in paddle_ocr_result[0]]))

        # Extract the words and bounding boxes from the OCR results
        words = []
        boxes = []
        scores = []
        for line in paddle_ocr_result:
            for bbox in line:
                words.append(bbox[1][0])
                scores.append(bbox[1][1])
                boxes.append(bbox[0])
        
        output_image = cv2.rectangle(image,(x1,y1),(x2,y2),(0,0,255),2)
        cv2.putText(output_image,str(words),(x1,y1),cv2.FONT_HERSHEY_PLAIN,2,(255,0,255),3)

    return words, boxes, scores, output_image    

# %%
def extract_License_Number_EasyOCR(image_path,conf_threshold ):
    image = cv2.imread(image_path)
    # image = cv2.resize(image,(1020,800))
    results = get_LicencePlate(image ,conf_threshold=0.1  )
    easyOCR_output = []
    paddle_output = []

    # print("results************************", results)
    for index in results:
        x1 = index[0]
        y1 = index[1]
        x2 = index[2]
        y2 = index[3]
        class_name=(index[4])

        cropped_img = image[y1:y2, x1:x2]
        processed_img =  cropped_img

        easyocr_reader = easyocr.Reader(['en'], verbose=False)
        easyocr_result = easyocr_reader.readtext(processed_img)

        easyOCR_output.append(easyocr_result)     

        print('\nEasyOCR:')
        print(''.join([text[1] + ' ' for text in easyocr_result]))

        # Extract the words and bounding boxes from the OCR results
        words = []
        boxes = []
        scores = []
        for line in easyocr_result:
            words.append(line[1])
            scores.append(line[2])
            boxes.append(line[0])
        output_image = cv2.rectangle(image,(x1,y1),(x2,y2),(0,0,255),2)
        cv2.putText(output_image,str(words),(x1,y1),cv2.FONT_HERSHEY_PLAIN,1,(255,0,255),2)
  
    return words, boxes, scores, output_image

# %%
def createDataframe(words, boxes, scores):
    df = pd.DataFrame(list(zip(words, boxes, scores)), columns=['words', 'boxes', 'scores'])
    return df

# %%
def initiate_Extract(image_path, conf_threshold: gr.inputs.Slider = 0.10, ocr_type="PaddleOCR"):

    if ocr_type == "PaddleOCR":
        words, boxes, scores, output_img = extract_License_Number_PaddleOCR(image_path, conf_threshold)
    elif ocr_type == "EasyOCR":
        words, boxes, scores, output_img = extract_License_Number_EasyOCR(image_path, conf_threshold)
    else:
        words, boxes, scores, output_img = extract_License_Number_PaddleOCR(image_path ,conf_threshold )

    dataframe = createDataframe(words, boxes, scores)
    return cv2.cvtColor(output_img, cv2.COLOR_BGR2RGB), dataframe


# %%
import numpy as np
title = "License Plate Number Recognition using YOLO & Paddle OCR - EasyOCR"
description = ""

css = """.output_image, .input_image {height: 600px !important}"""
examples = [['truck1.jpg'],['c1.jpg']]
iface = gr.Interface(fn=initiate_Extract,
                     inputs=[
                        gr.inputs.Image(type="filepath", label="Input Image"),
                        gr.inputs.Slider(minimum=0.0, maximum=1.0, default=0.10, step=0.05, label="Confidence Threshold"),
                        gr.inputs.Dropdown(label="Select the OCR",default="PaddleOCR", choices=["PaddleOCR", "EasyOCR"]),
                     ],
                     outputs=[gr.outputs.Image(type="pil", label="annotated image"),"dataframe"] ,
                     title=title,
                     description=description,
                     examples=examples,
                     css=css,
                     analytics_enabled = True, enable_queue=True)

iface.launch(inline=False , debug=True)

# %%


# %%