rgd_client - Resonant GeoDataClient

Python包rgd_client是针对Resonant GeoData API的强类型、易于使用和可扩展的Python客户端。

安装

要安装核心客户端

pip install rgd-client

要使用其他核心模块或插件，安装相应的客户端包。例如，影像客户端插件可以通过以下方式安装：

pip install rgd-imagery-client

通过插件添加的所有功能都在由该插件定义的名称下命名空间。对于影像客户端插件，这是imagery，因此可以通过client.imagery.*访问此插件的所有功能。以下是一些示例。

使用方法

搜索并显示结果

import json
import matplotlib.pyplot as plt
import numpy as np

from rgd_client import create_rgd_client

def plot_geojson(gjs, *args, **kwargs):
    points = np.array(gjs['coordinates'])
    if points.ndim == 3:
        points = points[0]
    if points.ndim == 1:
        points = points.reshape((1, points.size, ))
    return plt.plot(points[:,0], points[:,1], *args, **kwargs)

client = create_rgd_client(username='username', password='password')
bbox = {
    "type":"Polygon",
    "coordinates":[
        [
            [-105.45091240368326,39.626245373878696],
            [-105.45091240368326,39.929904289147274],
            [-104.88775649170178,39.929904289147274],
            [-104.88775649170178,39.626245373878696],
            [-105.45091240368326,39.626245373878696]
        ]
    ]
}

q = client.rgd.search(query=json.dumps(bbox), predicate='intersects')

for s in q:
    print(s['subentry_name'])

plot_geojson(bbox, 'k--', label='Search Region')

for s in q:
    plot_geojson(s['footprint'], label=s['subentry_name'])

plt.legend()
plt.title(f'Count: {len(q)}')

检查栅格

预览栅格缩略图

import imageio
from io import BytesIO

raster = client.imagery.get_raster(q[0])
plot_geojson(bbox, 'k--')
plot_geojson(raster['outline'], 'r')
load_image = lambda imbytes: imageio.imread(BytesIO(imbytes))

count = len(raster['parent_raster']['image_set']['images'])
for i in range(count):
    thumb_bytes = client.imagery.download_raster_thumbnail(q[0], band=i)
    thumb = load_image(thumb_bytes)
    plt.subplot(1, count, i+1)
    plt.imshow(thumb)

plt.tight_layout()
plt.show()

下载栅格

下载整个栅格图像集

import rasterio
from rasterio.plot import show

paths = client.imagery.download_raster(q[0])
rasters = [rasterio.open(im) for im in paths.images]
for i, src in enumerate(rasters):
    plt.subplot(1, len(rasters), i+1)
    ax = plt.gca()
    show(src, ax=ax)
plt.tight_layout()
plt.show()

STAC Item支持

Python客户端有一个专门针对栅格数据的搜索端点，该端点将搜索结果中的每个记录作为STAC Item返回。

q = client.imagery.search_raster_stac(query=json.dumps(bbox), predicate='intersects')

print(q[0])  # view result as STAC Item

# Download using the search result
paths = client.imagery.download_raster(q[0])
print(paths)

我们还可以上传STAC Item格式的新的数据。在这里，我们简单地返回相同的STAC Item JSON，实际上不会做任何事情，因为RGD已经识别出这些文件已经以栅格的形式存在。

client.imagery.create_raster_stac(q[0])

请注意，STAC Item中的资产必须已经上传到具有s3://或https:// URL的云存储提供者。此外，图像必须在每个资产上具有data标签。例如。

{
    ... # other STAC Item fields
    'assets': {
        'image-15030': {
            'href': 'http://storage.googleapis.com/gcp-public-data-sentinel-2/tiles/17/S/MS/S2A_MSIL1C_20210302T161201_N0209_R140_T17SMS_20210302T200521.SAFE/GRANULE/L1C_T17SMS_A029738_20210302T161751/IMG_DATA/T17SMS_20210302T161201_B01.jp2',
            'title': 'GRANULE/L1C_T17SMS_A029738_20210302T161751/IMG_DATA/T17SMS_20210302T161201_B01.jp2',
            'eo:bands': [{'name': 'B1'}],
            'roles': ['data'],
        },
        'image-15041': {
            'href': 'http://storage.googleapis.com/gcp-public-data-sentinel-2/tiles/17/S/MS/S2A_MSIL1C_20210302T161201_N0209_R140_T17SMS_20210302T200521.SAFE/GRANULE/L1C_T17SMS_A029738_20210302T161751/IMG_DATA/T17SMS_20210302T161201_B02.jp2',
            'title': 'GRANULE/L1C_T17SMS_A029738_20210302T161751/IMG_DATA/T17SMS_20210302T161201_B02.jp2',
            'eo:bands': [{'name': 'B1'}],
            'roles': ['data'],
        },
        ...  # ancillary files can lack a role but we like to see `metadata` used.
        'ancillary-30687': {
            'href': 'http://storage.googleapis.com/gcp-public-data-sentinel-2/tiles/17/S/MS/S2A_MSIL1C_20210302T161201_N0209_R140_T17SMS_20210302T200521.SAFE/GRANULE/L1C_T17SMS_A029738_20210302T161751/QI_DATA/MSK_TECQUA_B03.gml',
            'title': 'GRANULE/L1C_T17SMS_A029738_20210302T161751/QI_DATA/MSK_TECQUA_B03.gml',
            'roles': ['metadata'],
        },
    }
}

插件开发

有关如何为 rgd_client 开发插件的说明，请参阅 PLUGINS.md。