After tweaking the HSV values i'm able to achieve this result, now moving ahead i tried the canny edge detection but it doesn't produce any satisfactory result, My goal is to measure the size of plant. Any idea how do move ahead from this?
#Raspberry pi 3B+, debian Strect, kernel 4.19.5 v7
#OpenCV 4.0-pre
#Date 1st December, 2018
import cv2
import numpy as np
if __name__ == '__main__' :
# Read image
im = cv2.imread("leaf.jpg")
# Select ROI
fromCenter = False
r = cv2.selectROI(im, fromCenter )
# Crop image
imCrop = im[int(r[1]):int(r[1]+r[3]), int(r[0]):int(r[0]+r[2])]
cv2.imwrite('leaf2.jpg', imCrop)
# Display cropped image
cv2.imshow("Image", imCrop)
cv2.waitKey(0)
Drag rectangle to specify area. Then press space-bar to see second window.
thanks for the response @supra56, I think measuring the size of plant in pixels would be good idea. like counting the number of pixels Black & blue would give a clear indication of size of plant.
Asked: 2018-11-28 23:36:04 -0600
Seen: 1,269 times
Last updated: Dec 01 '18
you have a mask now, do findContours, than get the biggest one by minAreaRect or contourArea.