Description

I am trying to convert a numpy array and then, apply a color map to a 2d numpy array that’s filled with float values between 0 – 1. I have written this piece of code for that:

for i in range(arr.shape[0]):

    for j in range(arr.shape[1]):

        arr[i,j] = arr[i,j] * 255
        arr[i,j] = int(round(arr[i,j]))
        
arr.astype(np.uint8)
colorMapImage = cv2.cvtColor(arr,cv2.COLOR_GRAY2BGR)
colorMapImage = cv2.applyColorMap(colorMapImage, cv2.COLORMAP_VIRIDIS)

But, when I try to run the code I see the following error:

[ERROR] [1719390004.370940]: Error processing image: OpenCV(4.2.0) /home/myPC/opencv_build/opencv/modules/imgproc/src/colormap.cpp:714: error: (-5:Bad argument) cv::ColorMap only supports source images of type CV_8UC1 or CV_8UC3 in function 'operator()'

Notes

1. The numpy array was normalized to interval 0 – 1 before the conversion to OpenCV image.

   MEDevel.com: Open-source for Healthcare and Education

   Collecting and validating open-source software for healthcare, education, enterprise, development, medical imaging, medical records, and digital pathology.

   Visit Medevel

2. Shape of arr is (300, 300)

3. When I print colorMapImage[0,0] just before applying color map, It prints: [5. 5. 5.]

4. When I print out arr.dtype(), I see the following error:

   Error processing image: 'numpy.dtype[float32]' object is not callable

5. When I print types of both arr and colorMapImage with print(str(type())), it prints, <class ‘numpy.ndarray’> for both.

>Solution :

The arr numpy array is not being correctly converted to an 8-bit unsigned integer (np.uint8) before applying the color map. The conversion to np.uint8 is attempted, but the result is not stored back into arr. There also seems to be a minor mistake in how you’re trying to print the data type of arr.

import numpy as np
import cv2

arr = (arr * 255).round().astype(np.uint8)
colorMapImage = cv2.cvtColor(arr, cv2.COLOR_GRAY2BGR)
colorMapImage = cv2.applyColorMap(colorMapImage, cv2.COLORMAP_VIRIDIS)
print(arr.dtype)
print(colorMapImage[0, 0])