Extract car license plate number in image

Question

I want to find the car plate number to search in a database. Since Saudi plates are different, I face this problem

The result of the code

My current approach is to search for the cross in openCV using edge detection. How can I found the cross and take the below character (using container and edge detection)?

import numpy as np
import pytesseract
from PIL import Image
import cv2
import imutils
import matplotlib.pyplot as plt
import numpy as np

img = cv2.imread('M4.png')
img = cv2.resize(img, (820,680) )
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) #convert to grey scale
gray =  cv2.blur(gray, (3,3))#Blur to reduce noise
edged = cv2.Canny(gray, 10, 100) #Perform Edge detection
# find contours in the edged image, keep only the largest
# ones, and initialize our screen contour
cnts = cv2.findContours(edged.copy(), cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
cnts = imutils.grab_contours(cnts)
cnts = sorted(cnts, key = cv2.contourArea, reverse = True)[:10]
screenCnt = None

# loop over our contours
for c in cnts:
    # approximate the contour
    peri = cv2.arcLength(c, True)
    approx = cv2.approxPolyDP(c, 0.1 * peri, True)
    # if our approximated contour has four points, then
    # we can assume that we have found our screen
    if len(approx) == 4:
        screenCnt = approx
        break
if screenCnt is None:
    detected = 0
    print "No contour detected"
else:
    detected = 1

if detected == 1:
    cv2.drawContours(img, [screenCnt], -1, (0, 255, 0), 3)
# Masking the part other than the number plate
imgs = img
mask = np.zeros(gray.shape,np.uint8)
new_image = cv2.drawContours(mask,[screenCnt],0,255,-1,)
new_image = cv2.bitwise_and(imgs,imgs,mask=mask)

# Now crop
(x, y) = np.where(mask == 255)
(topx, topy) = (np.min(x), np.min(y))
(bottomx, bottomy) = (np.max(x), np.max(y))
Cropped = gray[topx:bottomx+1, topy:bottomy+1]

#Read the number plate
text = pytesseract.image_to_string(Cropped, config='--psm 11')
print("Detected Number is:",text)
plt.title(text)
plt.subplot(1,4,1),plt.imshow(img,cmap = 'gray')
plt.title('Original'), plt.xticks([]), plt.yticks([])
plt.subplot(1,4,2),plt.imshow(gray,cmap = 'gray')
plt.title('gray'), plt.xticks([]), plt.yticks([])
plt.subplot(1,4,3),plt.imshow(Cropped,cmap = 'gray')
plt.title('Cropped'), plt.xticks([]), plt.yticks([])
plt.subplot(1,4,4),plt.imshow(edged,cmap = 'gray')
plt.title('edged'), plt.xticks([]), plt.yticks([])
plt.show()

#check data base

#recoed the entre

cv2.waitKey(0)
cv2.destroyAllWindows()

Thanks for your help

Answer 1

Here's an approach:

• Convert image to grayscale and Gaussian blur
• Otsu's threshold to get a binary image
• Find contours and sort contours from left-to-right to maintain order
• Iterate through contours and filter for the bottom two rectangles
• Extract ROI and OCR

---

After converting to grayscale and Gaussian blurring, we Otsu's threshold to get a binary image. We find contours then sort the contours using imutils.contours.sort_contours() with the left-to-right parameter. This step keeps the contours in order. From here we iterate through the contours and perform contour filtering using these three filtering conditions:

• The contour must be larger than some specified threshold area (3000)
• The width must be larger than the height
• The center of each ROI must be in the bottom half of the image. We find the center of each contour and compare it to where it is located on the image.

If a ROI passes these filtering conditions, we extract the ROI using numpy slicing and then throw it into Pytesseract. Here's the detected ROIs that pass the filter highlighted in green

Since we already have the bounding box, we extract each ROI

We throw each individual ROI into Pytesseract one at a time to construct our license plate string. Here's the result

License plate: 430SRU

Code

import cv2
import pytesseract
from imutils import contours

pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"

image = cv2.imread('1.png')
height, width, _ = image.shape
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
blur = cv2.GaussianBlur(gray, (5,5), 0)
thresh = cv2.threshold(blur, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)[1]
cnts = cv2.findContours(thresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
cnts, _ = contours.sort_contours(cnts, method="left-to-right")

plate = ""
for c in cnts:
    area = cv2.contourArea(c)
    x,y,w,h = cv2.boundingRect(c)
    center_y = y + h/2
    if area > 3000 and (w > h) and center_y > height/2:
        ROI = image[y:y+h, x:x+w]
        data = pytesseract.image_to_string(ROI, lang='eng', config='--psm 6')
        plate += data

print('License plate:', plate)