class Extractor(object):
'''
Stores and manipulates the input image to extract the Sudoku puzzle
all the way to the cells
'''
def __init__(self, path):
self.helpers = Helpers() # Image helpers
self.image = self.loadImage(path)
self.preprocess()
#self.helpers.show(self.image, 'After Preprocessing')
sudoku = self.cropSudoku()
#self.helpers.show(sudoku, 'After Cropping out grid')
sudoku = self.straighten(sudoku)
#self.helpers.show(sudoku, 'Final Sudoku grid')
self.cells = Cells(sudoku).cells
def loadImage(self, path):
color_img = cv2.imread(path)
if color_img is None:
raise IOError('Image not loaded')
print 'Image loaded.'
return color_img
def preprocess(self):
print 'Preprocessing...',
self.image = cv2.cvtColor(self.image, cv2.COLOR_BGR2GRAY)
self.image = self.helpers.thresholdify(self.image)
kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (2, 2))
self.image = cv2.morphologyEx(self.image, cv2.MORPH_CLOSE, kernel)
print 'done.'
def cropSudoku(self):
print 'Cropping out Sudoku...',
contour = self.helpers.largestContour(self.image.copy())
sudoku = self.helpers.cut_out_sudoku_puzzle(self.image.copy(), contour)
print 'done.'
return sudoku
def straighten(self, sudoku):
print 'Straightening image...',
largest = self.helpers.largest4SideContour(sudoku.copy())
app = self.helpers.approx(largest)
corners = self.helpers.get_rectangle_corners(app)
sudoku = self.helpers.warp_perspective(corners, sudoku)
print 'done.'
return sudoku
class Extractor(object):
'''
Stores and manipulates the input image to extract the Sudoku puzzle
all the way to the cells
'''
digits = []
def __init__(self, path):
self.helpers = Helpers() # Image helpers
if (path == 'error1'):
self.errorimg = self.loadImage('input.jpeg')
W = self.errorimg.shape[0]
H = self.errorimg.shape[1]
cv2.putText(self.errorimg, "Grid not Detected",
(0, int(H / 2)), cv2.FONT_HERSHEY_SIMPLEX,
int(1 * (W / 400)), (0, 0, 255), 3)
self.helpers.show(self.errorimg, 'Final Sudoku grid')
return
if (path == 'error2'):
self.errorimg = self.loadImage('input.jpeg')
W = self.errorimg.shape[0]
H = self.errorimg.shape[1]
cv2.putText(self.errorimg, "Could not Solve",
(0, int(H / 2)), cv2.FONT_HERSHEY_SIMPLEX,
int(1 * (W / 400)), (0, 0, 255), 3)
self.helpers.show(self.errorimg, 'Final Sudoku grid')
return
self.image = self.loadImage(path)
self.preprocess()
#self.helpers.show(self.image, 'After Preprocessing')
self.sudoku = self.cropSudoku()
#self.helpers.show(sudoku, 'After Cropping out grid')
self.sudoku = self.straighten(self.sudoku)
#self.helpers.show(self.sudoku, 'Final Sudoku grid')
self.digits = Cells(self.sudoku).cells
def displayDigets(self, sudoku, inputArray, digets):
sudokuDisplay = cv2.cvtColor(sudoku, cv2.COLOR_GRAY2RGB)
sudokuDisplay = cv2.bitwise_not(sudokuDisplay)
W = sudoku.shape[0]
H = sudoku.shape[1]
W = W // 9
H = H // 9
for i in range(0, 81):
if (not (digets[i] == 0) and inputArray[i] == 0):
cv2.putText(
sudokuDisplay, str(digets[i]),
(int(W * (i % 9) + W / 4), int(H * (i // 9) + 3 * H / 4)),
cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 0, 255), 3)
self.helpers.show(sudokuDisplay, 'Final Sudoku grid')
return sudoku
def loadImage(self, path):
color_img = cv2.imread(path)
if color_img is None:
raise IOError('Image not loaded')
print('Image loaded.')
return color_img
def preprocess(self):
print('Preprocessing...', end=' ')
self.image = cv2.cvtColor(self.image, cv2.COLOR_BGR2GRAY)
self.image = self.helpers.thresholdify(self.image)
kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (2, 2))
self.image = cv2.morphologyEx(self.image, cv2.MORPH_CLOSE, kernel)
print('done.')
def cropSudoku(self):
print('Cropping out Sudoku...', end=' ')
contour = self.helpers.largestContour(self.image.copy())
sudoku = self.helpers.cut_out_sudoku_puzzle(self.image.copy(), contour)
print('done.')
return sudoku
def straighten(self, sudoku):
print('Straightening image...', end=' ')
largest = self.helpers.largest4SideContour(sudoku.copy())
app = self.helpers.approx(largest)
corners = self.helpers.get_rectangle_corners(app)
sudoku = self.helpers.warp_perspective(corners, sudoku)
print('done.')
return sudoku
