OpenCV question about thresholding

Technical Programming
1

We are using OpenCV (python) to threshold an image. I saw team 254’s presentation, which said to use HSV/HSL. We are using HSV. However, I believe that OpenCV is still expecting a BGR format. How do I have it take in an HSV/HSL formatted color?

Here is our code:

import numpy as np
import cv2
from networktables import NetworkTable as nettab
import logging

logging.basicConfig(level=logging.DEBUG)

#H: 0 - 360
#S: 0 - 255
#V: 0 - 255

lower_thresh = np.array([60, 150, 150])
upper_thresh = np.array([180, 255, 255])

#lower_thresh = np.array([0, 100, 0])
#upper_thresh = np.array([50, 255, 50])

lower_thresh[0] /= 2
upper_thresh[0] /= 2

nettab.setIPAddress("10.51.13.2")
nettab.setClientMode()
nettab.initialize(server="10.51.13.83")
table = nettab.getTable("vision")

print(nettab.isConnected())

cap = cv2.VideoCapture(0) #1280 x 720

while (True):
    # Capture frame-by-frame
    ret, frame = cap.read()  # frame = None, which is causing the run time error

    # Our operations on the frame come here
    new = cv2.inRange(frame, lower_thresh, upper_thresh)

    #Display the resulting frame
    cv2.imshow('BW', new)
    cv2.imshow('Color', frame)

    table.putBoolean("test", True)

    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

# When everything done, release the capture
cap.release()
cv2.destroyAllWindows()
2

You will need to convert the BGR to HSV.

#Convert the image from BGR(RGB) to HSV
hsvImage = cv2.cvtColor( Raw, cv2.COLOR_BGR2HSV)

Once converted to HSV, thresholding is simple.
We keep our upper and lower threshold values in a “calibration” file.

CalFile = open ('Calibration').read().splitlines()
uh = int(CalFile[0])
lh = int(CalFile[1])
us = int(CalFile[2])
ls = int(CalFile[3])
uv = int(CalFile[4])
lv = int(CalFile[5])
er = int(CalFile[6])
dl = int(CalFile[7])
ap = int(CalFile[8])

Our thresholding command looks like this.

## Threshold HSV Image to find specific color
binImage = cv2.inRange(hsvImage, (lh, ls, lv), (uh, us, uv))
3

Could you link to the 254 presentation?