opencv - Segmentation problem for tomato leaf images in PlantVillage Dataset

Question

I am trying to do segmentation of leaf images of tomato crops. I want to convert images like following image

to following image with black background

I have reference this code from Github

but it does not do well on this problem, It does something like this

Can anyone suggest me a way to do it ?

Answer 1

The image is separable using the HSV-colorspace. The background has little saturation, so thresholding the saturation removes the gray.

Result:

Code:

import numpy as np 
import cv2
# load image
image = cv2.imread('leaf.jpg')
# create hsv
hsv = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)
 # set lower and upper color limits
low_val = (0,60,0)
high_val = (179,255,255)
# Threshold the HSV image 
mask = cv2.inRange(hsv, low_val,high_val)
# remove noise
mask = cv2.morphologyEx(mask, cv2.MORPH_CLOSE, kernel=np.ones((8,8),dtype=np.uint8))
# apply mask to original image
result = cv2.bitwise_and(image, image,mask=mask)

#show image
cv2.imshow("Result", result)
cv2.imshow("Mask", mask)
cv2.imshow("Image", image)

cv2.waitKey(0)
cv2.destroyAllWindows()