imgSeg.py

def imgSeg(path):
	
	from urllib import request
	import cv2
	import numpy as np

	# URL to image
	resp = request.urlopen(path)
	img = np.asarray(bytearray(resp.read()), dtype="uint8")
	img = cv2.imdecode(img, cv2.IMREAD_COLOR)

	gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
	laplacian = cv2.Laplacian(gray, cv2.CV_8UC1)
	sobel = cv2.Sobel(gray, cv2.CV_8UC1, 0, 1, ksize= 3)

	# Remove some small noise if any.
	dilate = cv2.dilate(sobel,None)
	erode = cv2.erode(dilate,None)

	ret,thresh1 = cv2.threshold(erode,127,255,cv2.THRESH_BINARY)

	# Find contours with cv2.RETR_CCOMP
	contours, hierarchy = cv2.findContours(erode, cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)[-2:]

	# Get width, height of origin image
	height, width, channels = img.shape
	# Temp criteria of width
	tempWidth = width - 100

	count = 0
	tempY = height
	
	for i,cnt in enumerate(contours):
	    # Check if it is an external contour and its area is more than 100
	    if hierarchy[0,i,3] == -1 : 
	    	x,y,w,h = cv2.boundingRect(cnt)
	    	if(w > tempWidth):
                        cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,0),2)

                        m = cv2.moments(cnt)
                        tempHeight = height -( height- tempY )

                        if( tempHeight -y > 100 ) :
                            count += 1
                            cv2.imwrite("output/Img"+str(count)+".jpg", img[y:tempHeight, 0:0+width])
                            tempY = y
	if(tempY > 10):
		count += 1
		cv2.imwrite("output/Img"+str(count)+".jpg", img[0:tempY, 0:0+width])
	print(count)        
	#cv2.imwrite("output/Img"+str(count)+".jpg", img[0:tempY, 0:0+width])
	#cv2.imshow('img', temp)
	#cv2.waitKey(0)
	#cv2.destroyAllWindows()

imgSeg('http://gi.esmplus.com/orgastore/img/hobak_total.jpg')