Some final comments and added some sources that can be found in my notebook as well

board_detector.py

@@ -95,6 +95,8 @@ def find_lines(img):
     return sorted_vertical, sorted_horizontal
 
 def convert_to_cartesian(lines):
+    # used for HoughLines
+    # www.geeksforgeeks.org/line-detection-python-opencv-houghline-method/
     line_points = []
     if lines is not None:
         for line in lines:
@@ -145,6 +147,10 @@ def filter_lines(lines, min_distance):
     return filtered_lines
 
 def find_board(img):
+    # https://opencvpython.blogspot.com/2012/06/sudoku-solver-part-2.html
+    # https://stackoverflow.com/questions/10196198/how-to-remove-convexity-defects-in-a-sudoku-square/11366549#11366549
+    # ^ two sources that were very useful for the board detection code
+
     vertical_lines, horizontal_lines = find_lines(img)
 
     # create bitmasks for vert and horiz so we can get lines and intersections
@@ -280,7 +286,7 @@ def warp_board(img, corners_sorted, intersection):
 
     return warped_img, sorted_warped_points
 
-prev_filled_contour_mask = None
+prev_filled_contour_mask = None # probably should have stored this in game.py instead
 
 def find_pieces(warped_img, prev_warped_img, sorted_warped_points, reader, img_idx):
     global prev_filled_contour_mask
@@ -485,6 +491,7 @@ def find_pieces(warped_img, prev_warped_img, sorted_warped_points, reader, img_i
                 
                 if len(box_contours_sorted) != 0:
                     # straighten the letter based on its contours' bounding box
+                    # https://stackoverflow.com/questions/11627362/how-to-straighten-a-rotated-rectangle-area-of-an-image-using-opencv-in-python <- this helped a lot
                     box_contour = box_contours_sorted[0]
                     drawn_box_img = cv2.cvtColor(box_img, cv2.COLOR_GRAY2BGR)
                     cv2.drawContours(drawn_box_img, [box_contour], -1, (255, 255, 0), 1)
@@ -547,7 +554,7 @@ def find_pieces(warped_img, prev_warped_img, sorted_warped_points, reader, img_i
                     img_to_read_bigger = np.zeros((cols + 40, rows + 40))
                     img_to_read_bigger = img_to_read_bigger.astype(np.uint8)
 
-                    # put img_to_read in center of img_to_read_bigger
+                    # put img_to_read in center of img_to_read_bigger (probably should have made this section a separate function)
                     start_row = int((img_to_read_bigger.shape[0] - img_to_read.shape[0]) // 2)
                     start_col = int((img_to_read_bigger.shape[1] - img_to_read.shape[1]) // 2)
                     end_row = start_row + img_to_read.shape[0]