HSV2BGR conversion fails in Python OpenCV script

Question

My script is supposed to take a greyscale image and map the values to hues.

#!/usr/bin/env python

import cv2
import numpy

infile = cv2.imread('Lenna.png')

infile = infile[:,:,0]

hues = (numpy.array(infile)/255.)*179

outimageHSV = numpy.array([[[b,255,255] for b in a] for a in hues]).astype(int)

outimageBGR = cv2.cvtColor(outimageHSV, cv2.COLOR_HSV2BGR)

cv2.imshow('dst_rt', outimageBGR)
cv2.waitKey(0)
cv2.destroyAllWindows()

It fails on the line with cvtColor and I get this error:

OpenCV Error: Assertion failed (depth == CV_8U || depth == CV_16U || depth == CV_32F) in cvtColor, file /tmp/opencv20150506-38415-u2kidu/opencv-2.4.11/modules/imgproc/src/color.cpp, line 3644
Traceback (most recent call last):
  File "luma2hue.py", line 16, in <module>
    outimageBGR = cv2.cvtColor(outimageHSV, cv2.COLOR_HSV2BGR)
cv2.error: /tmp/opencv20150506-38415-u2kidu/opencv-2.4.11/modules/imgproc/src/color.cpp:3644: error: (-215) depth == CV_8U || depth == CV_16U || depth == CV_32F in function cvtColor

Do I need to do something else to my outimageHSV array to make it ready for cvtColor?

Answer 1

The error message implies that cv2.cvtColor expects an image with a (color) depth of 8 or 16 bit unsigned int (8U, 16U) or 32 bit float (32F). Try changing astype(int) to astype(numpy.uint8)

Answer 2

outimageHSV needs to be casted as uint8.

import numpy as np

outimageHSV = np.uint8(outimageHSV)
outimageBGR = cv2.cvtColor(outimageHSV, cv2.COLOR_HSV2BGR)