Reputation: 117
I am doing some prediction of depth. So I used Colormap both in Opencv and Matplotlib on the predicted image. However, it performed well in Matplotlib but not in OpenCV(colormap_jet).
How can I solve this problem in Opencv? Because I want to use Opencv in real-time. Matplotlib is too slow in real-time.
Upvotes: 0
Views: 2066
Reputation: 18321
Use matplotlib.pylab.cm
to colorize the image.
#!/usr/bin/python3
# 2017.12.28 16:26:26 CST
import matplotlib.pyplot as plt
from matplotlib.pylab import cm
import numpy as np
import cv2
## use matplot jet for opencv
def colorize(img):
gray = None
if img.ndim == 2:
gray = img.copy()
if len(img.shape) == 3:
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
canvas = np.uint8(cm.jet(gray)*255)
canvas = cv2.cvtColor(canvas, cv2.COLOR_RGBA2BGR)
return canvas
## process
img = cv2.imread("test.png")
res = colorize(img)
cv2.imwrite("res.png", res)
Example 1:
Generate the colorized the image (as the preview code).
Example 2:
Colorize the same data in Matplotlib
(in jet
) and OpenCV
(using my function colorize
).
Upvotes: 1