下面是在Python/OpenCV中实现这一点的一种方法。

• 读取输入
• 列出原点的Numpy数组
• 在数组的输入副本上绘制多边形
• 扩大点的数组以围绕图像的顶部生成一个多边形
• 在黑色背景上绘制一个白色填充多边形作为mask1
• 使用mask1涂黑输入图像的底部并保留顶部
• 扩大点的数组以围绕图像底部生成一个多边形
• 在黑色背景上绘制一个白色填充多边形作为mask2
• 使用mask2涂黑输入图像的顶部并保留底部
• 保存结果

输入：

import cv2
import numpy as np

# read the input
img = cv2.imread('fish.png')
h, w = img.shape[:2]

# create black image like input
black = np.zeros_like(img)

# define original point
points = np.array([[954,88],[905,97],[855,107],[805,114],[755,125],[705,134],[655,139],[605,141],[555,139],[505,134],[455,139],[405,146],[355,146],[305,144],[255,145],[205,125],[155,120],[105,114],[55,104],[1,156]])

# draw points on input
img_pts = img.copy()
cv2.polylines(img_pts, [points], False, (0,0,255), 2)

# augment original polygon array for top region
points1 = np.array([[0,0],[w-1,0],[w-1,88],[954,88],[905,97],[855,107],[805,114],[755,125],[705,134],[655,139],[605,141],[555,139],[505,134],[455,139],[405,146],[355,146],[305,144],[255,145],[205,125],[155,120],[105,114],[55,104],[1,156],[0,156]])

# draw white filled closed polygon on black background
mask1 = black.copy()
cv2.fillPoly(mask1, [points1], (255,255,255))

# Use mask to select top part of image
result1 = np.where(mask1==255, img, black)

# augment polygon array for bottom region
points2 = np.array([[0,h-1],[w-1,h-1],[w-1,88],[954,88],[905,97],[855,107],[805,114],[755,125],[705,134],[655,139],[605,141],[555,139],[505,134],[455,139],[405,146],[355,146],[305,144],[255,145],[205,125],[155,120],[105,114],[55,104],[1,156],[0,156]])

# draw white filled closed polygon on black background
mask2 = black.copy()
cv2.fillPoly(mask2, [points2], (255,255,255))

# Use mask to select part of image
result2 = np.where(mask2==255, img, black)

# save results
cv2.imwrite('fish_with_points.jpg', img_pts)
cv2.imwrite('fish_top_mask.jpg', mask1)
cv2.imwrite('fish_top_masked.jpg', result1)
cv2.imwrite('fish_bottom_mask.jpg', mask2)
cv2.imwrite('fish_bottom_masked.jpg', result2)

# show results
cv2.imshow('img_points',img_pts)
cv2.imshow('mask1',mask1)
cv2.imshow('result1',result1)
cv2.imshow('mask2',mask2)
cv2.imshow('result2',result2)
cv2.waitKey(0)

输入时的原点折线：

面罩1：

结果1：

掩模2：

结果2：