Breadth First Search optimisation of recursive connected components labeling algorithm

labeling.py

import Image
import Queue


def select_label(used):
    """
    :type used: set
    :return: int
    """
    prev = 0
    for label in used:
        if label-1 is not prev:
            used.add(prev+1)
            return prev+1
        else:
            prev = label
    if prev+1 not in used:
        used.add(prev+1)
    
    return prev+1


def __fill(data, labels, size, q, args):
    i, label = args
    width, height = size
    col, row = i % width, int(i / height) + 1
    if data[i] is 0 or labels[i] is not 0:
        return False

    labels[i] = label
    if col > 0:
        q.put((i-1, label))
    if col < width - 1:
        q.put((i+1, label))
    if row > 0:
        q.put((i-width, label))
    if row < height - 1:
        q.put((i+width, label))

    return True


def recursive_labeling(bin_img):
    """
    :type bin_img: Image.Image
    :return: Image.Image
    """
    #Breadth First Search optimisation
    q = Queue.Queue()
    l_img = Image.new('L', bin_img.size)
    data = list(bin_img.getdata())
    labels = list(l_img.getdata())
    used_labels = set()
    l = select_label(used_labels)
    for i in xrange(len(data)-1):
        if __fill(data, labels, bin_img.size, q, (i, l)):
            l = select_label(used_labels)
        while not q.empty():
            args = q.get()
            __fill(data, labels, bin_img.size, q, args)

    l_img.putdata(labels)

    return l_img