Next, we carry out our initial search to find plate characters. First, we find characters roughly, and then find candidates based on specific criteria.
Let's start with the following line in our Notebook:
%pylab notebook
We can now execute our function cell for imports, utilities, and to load our libraries:
import cv2
import numpy as np
import pickle
def getmatchingchars(char_cands):
char_list = []
for char_cand in char_cands:
ch_matches = [] \n",
for matching_candidate in char_cands:
if matching_candidate == char_cand:
continue
chardistance = np.sqrt((abs(char_cand.x_cent - matching_candidate.x_cent) ** 2) +
(abs(char_cand.y_cent - matching_candidate.y_cent)**2))
x = float(abs(char_cand.x_cent - matching_candidate.x_cent))
y = float(abs(char_cand.y_cent - matching_candidate...