1.对于RGB三通道图片，直接用两层for循环的话，效率比较低
2.可以先将RGB图片转为灰度图片，再利用numpy.where的广播机制统计像素个数。这里有一个前提是提前知道与灰度图片的像素值相对应RGB颜色。
代码如下：
from PIL import Image
import numpy as np
import cv2
img_L = np.array(Image.open('test.png').convert("L"))
img_RGB = np.array(Image.open('test.png').convert("RGB"))
# temp = {}
# for i in range(img_L.shape[0]):
# for j in range(img_L.shape[1]):
# if not temp.get(int(img_L[i][j])):
# temp[int(img_L[i][j])] = list(img_RGB[i][j])
# print(temp)
#这里得到灰度像素值0对应(0,0,0),62对应(19,69,139)
color_0_0_0 = np.where(img_L == 0)[0].shape[0]
color_19_69_139 = np.where(img_L == 62)[0].shape[0]
pixel_sum = img_L.shape[0] * img_L.shape[1]
print("0_0_0 像素个数：{} 占比：%{}".format(color_0_0_0,color_0_0_0/pixel_sum*100))
print("19_69_139 像素个数：{} 占比：%{}".format(color_19_69_139,color_19_69_139/pixel_sum*100))
补充：OpenCV---如何统计图像的像素分布值个数（6）
import cv2 as cv
import matplotlib.pyplot as plt
import numpy as np
def statistics():
src = cv.imread("D:/matplotlib/0.jpg")
cv.imshow("q",src)
h,w,ch = np.shape(src)
gray = cv.cvtColor(src,cv.COLOR_BGR2GRAY)
cv.imshow("gray",gray)
hest = np.zeros([256],dtype = np.int32)
for row in range(h):
for col in range(w):
pv = gray[row,col]
hest[pv] +=1
plt.plot(hest,color = "r")
plt.xlim([0,256])
plt.show()
cv.waitKey(0)
cv.destroyAllWindows()
statistics()
像素分布统计图
import cv2 as cv
import matplotlib.pyplot as plt
import numpy as np
def statistics():
src = cv.imread("D:/matplotlib/0.jpg")
cv.imshow("q",src)
h,w,ch = np.shape(src)
#读取图像属性
gray = cv.cvtColor(src,cv.COLOR_BGR2GRAY)
#将图像转换成灰度图，
cv.imshow("gray",gray)
hest = np.zeros([256],dtype = np.int32)
#建立空白数组
for row in range(h):
for col in range(w):
pv = gray[row,col]
hest[pv] +=1
#统计不同像素值出现的频率
plt.plot(hest,color = "r")
plt.xlim([0,256])
plt.show()
#画出统计图
cv.waitKey(0)
cv.destroyAllWindows()
statistics()
以上为个人经验，希望能给大家一个参考，也希望大家多多支持。如有错误或未考虑完全的地方，望不吝赐教。
