Python 2.7 - CV2，Rasterio 收到错误 numpy.ndarray 对象不可调用

Question

我正在尝试用来自geojson的形状来掩盖图像。

我要做的第一件事是获取代表图像中经度，长坐标的像素。

当我获得这些信息时，我使用该坐标构建一个"多边形"，然后尝试裁剪图像，但出现此错误：

'numpy.ndarray' 对象不可调用

请注意，绝对路径是指.jp2文件。

geoms = [{'type':'Polygon', 'coordinates': [[(x1,y1),(x2,y2),(x3,y3),(x4,y4)]]}]
with rasterio.open(absolutePath) as src:
out_image,out_transform = mask(src,geoms,crop=True)
out_meta = src.meta.copy()

我也尝试使用第一个坐标作为第 5 个坐标来"密封"多边形，但没有任何变化。

我收到的唯一错误消息是这样的：

TypeErrorTraceback (most recent call last)
<ipython-input-13-796462428f9b> in <module>()
2 
3 with rasterio.open(absolutePath) as src:
----> 4     out_image,out_transform = mask(src,geoms,crop=True)
5 out_meta = src.meta.copy()
TypeError: 'numpy.ndarray' object is not callable

这是我的所有代码：

from snappy import ProductIO
from snappy import PixelPos, GeoPos
import numpy as np
import geojson
import cv2
import os
import rasterio
from rasterio.mask import mask
path = '/home/.../x.SAFE'
product = ProductIO.readProduct(path)
sg = product.getSceneGeoCoding()
pathGeoJson = '/home/.../x.geojson'
with open(pathGeoJson) as f:
gj = geojson.load(f)
features = gj['features'][0]
latlon1 = features.geometry.coordinates[0][0]
latlon2 = features.geometry.coordinates[0][1]
latlon3 = features.geometry.coordinates[0][2]
latlon4 = features.geometry.coordinates[0][3]
geoms = gj['features'][0].geometry
geoms
def LatLon_from_XY(ProductSceneGeoCoding, x, y):
geoPos = ProductSceneGeoCoding.getGeoPos(PixelPos(x,y),None)
lat = geoPos.getLat()
lon = geoPos.getLon()
return lat,lon
def XY_from_LatLon(ProductSceneGeoCoding, latitude, longitude):
pixelPos = ProductSceneGeoCoding.getPixelPos(GeoPos(latitude, longitude),None)
x = np.round(pixelPos.getX())
y = np.round(pixelPos.getY())
return x,y
x1,y1 = XY_from_LatLon(sg,latlon1[0], latlon1[1])
x2,y2 = XY_from_LatLon(sg,latlon2[0], latlon2[1])
x3,y3 = XY_from_LatLon(sg,latlon3[0], latlon3[1])
x4,y4 = XY_from_LatLon(sg,latlon4[0], latlon4[1])
print(x1,y1)
print(x2,y2)
print(x3,y3)
print(x4,y4)
pathToJP2 = '/home/.../IMG_DATA/'
arr = os.listdir(pathToJP2)
absolutePath = pathToJP2+arr[1]
arr[1]
print(absolutePath)
geoms = [{'type':'Polygon', 'coordinates': [[(x1,y1),(x2,y2),(x3,y3),(x4,y4)]]}]
with rasterio.open(absolutePath) as src:
out_image,out_transform = mask(src,geoms,crop=True)
out_meta = src.meta.copy()

Answer 1

您可以使用openCV进行裁剪甚至应用蒙版，这里回答得更好