为什么在Python中做透视变换之后输出的图片会变色？

Question

import cv2
import numpy as np
import matplotlib.pylab as plt


a = cv2.imread('11.jpg')
b = cv2.imread('3.jpg')

pts = np.float32([[0, 0], [0, 378], [378, 378], [378, 0]])

pts1 = np.float32([[0, 56.7], [0, 321.3], [326.8, 378], [326.8, 0]])
pts2 = np.float32([[0, 0], [0, 378], [326.8, 321.3], [326.8, 56.7]])

M = cv2.getPerspectiveTransform(pts, pts1)
N = cv2.getPerspectiveTransform(pts, pts2)

dst1 = cv2.warpPerspective(a, M, (378, 378))
dst2 = cv2.warpPerspective(b, N, (378, 378))

plt.figure()
plt.subplot(221)
plt.imshow(a)
plt.subplot(222)
plt.axis('off')#去掉坐标轴
plt.imshow(dst1)
plt.show()

Answer 1

cv2读取进来的是数字，通道顺序为BGR，因此使用plt显示结果颜色异常