rasterio.mask易错用法，掩膜之后是三维数组

def clip_veg(cities, raster_path, row_index):
    with rasterio.open(raster_path) as src:
        city = cities.iloc[row_index]
        city_geometry = city['geometry']
        out_image, out_transform = mask(src, [city_geometry], crop=True)
        out_meta = src.meta.copy()
        out_meta.update({
            "driver": "GTiff",
            "height": out_image.shape[1],
            "width": out_image.shape[2],
            "transform": out_transform
        })
        print(out_image.shape)
        return out_image, out_transform, out_meta

def veg_topoints(image, transform):
    rows, cols = np.where((image > 0) & (image < 255))
    x_coords = transform[0] + (cols + 0.5) * transform[1]
    y_coords = transform[3] + (rows + 0.5) * transform[5]
    geometries = [Point(x, y) for x, y in zip(x_coords, y_coords)]
    values = image[rows, cols]
    return gpd.GeoDataFrame({'raster_val': values, 'geometry': geometries}, crs='epsg:4326')

写了两个函数对数据依次操作，之后结果竟然报错，image是三维数组，查了一下rasterio.mask.mask的用法，裁剪之后是三维数组，就算原数据只有一个波段，最后也是（1，height, width），所以下一个函数np.where()部分就报错了。加入image=image[0]，变成二维数组就好了。