When we run our code using the best candidate code block, we get the following result:
if len(plate_candidates) > 0:
plate_candidates.sort(key = lambda plate_candidate:
len(plate_candidate.chars), reverse = True)
best_plate = plate_candidates[0]
print("License plate read: " + best_plate.chars + "\n")
License plate read: LTLDBENZ
Once we get our output, we can display our result using the following code:
figure()
imshow(best_plate.thesholded)
The displayed image will be as follows:
Although there is an extra character, we can see that our displayed image is very close to the plate characters. We can, check it with our other possible plate characters to get the closest result.
Let's try one more license plate, to check how our code works:
input_image = plt.imread('tests/p2.jpg') #use cv2.imread...