类型错误：使用 OpenCV 在 Python 中旋转图像时，参数“mat”需要 cv::UMat

Question

我试图使用 Python 旋转图像，并且我认为旋转图像的逻辑是正确的，但是每当我尝试使用 OpenCV 显示旋转后的图像时，我都会收到以下错误：

类型错误：参数 'mat' 需要 cv::UMat

我是Python新手，第一次使用OpenCV，所以对此没有太多想法。旋转图像的类型是numpy.ndarray，我认为 OpenCV 可以工作nd.array

这是我的代码：

import cv2
import numpy as np

def rotateToNinety(imageProcessing):
    imageProcessing = imageProcessing[::-1]
    rotatedImage = np.array([[[None for z in range(3)] for x in range(len(imageProcessing))]for y in range(len(imageProcessing[0]))])

    for i in range(len(imageProcessing[0])):
        for j in range(len(imageProcessing)):
            for k in range(3):
                rotatedImage[i][j][k] = imageProcessing[j][i][k]
    return rotatedImage


image = cv2.imread("download (2).jpg",1)

print("Original Image".center(50,"="))
print(image)

cv2.imshow("Original",image)
cv2.waitKey(0)

print(type(image))


rotated_90 = rotateToNinety(image)
print("Rotated Image".center(50,"="))
print(rotated_90)

print(type(rotated_90))
cv2.imshow("rotated",rotated_90)
cv2.waitKey(0)
cv2.destroyAllWindows()

Answer 1

在你的情况下尝试更换

def rotateToNinety(imageProcessing):
    imageProcessing = imageProcessing[::-1]
    rotatedImage = np.array([[[None for z in range(3)] for x in range(len(imageProcessing))]for y in range(len(imageProcessing[0]))])

    for i in range(len(imageProcessing[0])):
        for j in range(len(imageProcessing)):
            for k in range(3):
                rotatedImage[i][j][k] = imageProcessing[j][i][k]
    return rotatedImage

和

def rotateToNinety(imageProcessing):
    h,w,c = imageProcessing.shape

    rotatedImage = np.zeros((h,w,c), dtype=np.uint8)

    imageProcessing = imageProcessing[::-1]

    for i in range(h):
        for j in range(w):
            for k in range(3):
                rotatedImage[i][j][k] = imageProcessing[j][i][k]
    return rotatedImage

在您的情况下，您将列与行混合在一起。

哦，并尝试np.zeros((height, width, num_channels))尽可能多地用于创建空白图像。