代码之家 › 专栏 › 技术社区 › lucians

带opencv(python)的图像掩码

mask opencv python

lucians · 技术社区 · 8 年前

我编写了这个代码,它创建了2个遮罩。应用后,结果如下:

原始图像

产量

import cv2
import numpy as np

frame = cv2.imread('image.jpg')

h, w = frame.shape[:2]

upper_value = int(h / 10) * 5
lower_value = -(int(h / 10) * 3)

upper_mask = cv2.rectangle(frame, (0, 0), (w, upper_value), (0, 50, 255), -1)
lower_mask = cv2.rectangle(frame, (0, upper_value + int(h / 10) * 5), (w, upper_value + int(h / 10) * 2), (0, 50, 255), -1)

我知道代码,它一点都不好,但做它是工作。我该如何改进?

谢谢

1 回复 | 直到 8 年前

Jundiaius 8 年前

以下是一些建议:

import cv2
import numpy as np

frame = cv2.imread('image.jpg')

h, w = frame.shape[:2]
mask_color = (0, 50, 255) # isolate a repeating constant

# avoid setting a variable that is used only once, only if you REALLY need it to improve readability
# that's why `upper_value` and `lower_value` were removed. 
# BTW, `lower_value` was unused in your code.

upper_mask = cv2.rectangle(frame, (0, 0), (w, int(0.5 * h)), mask_color, -1)
lower_mask = cv2.rectangle(frame, (0, h), (w, int(0.7 * h)), mask_color, -1)

推荐文章

anfas2 · 如何使用MediaPipe在Python中检测到的地标上叠加自定义形状?

1 年前