raster-process

A raster data process utilities.

安装

npm i @sakitam-gis/raster-process -S

import { RasterProcess } from '@sakitam-gis/raster-process'

使用

const process = new RasterProcess();
process.use(task).use(task);

处理 GFS 格点数据

const path = require('path');

require('dotenv').config({
  path: path.resolve(process.cwd(), '.env'),
});

const { RasterProcess } =  require('../');

function run () {
  const rp = new RasterProcess();

  // const dataPath = path.resolve(__dirname, '../test/fixtures/gfs.t12z.pgrb2.0p25.grib');
  const dataPath = path.resolve(__dirname, './data/result/1.tiff');
  const tiffPath = path.resolve(__dirname, './data/result/gfs.t12z.pgrb2.0p25.tiff');
  const mercatorTiffPath = path.resolve(__dirname, './data/result/gfs.t12z.pgrb2.0p25-write-mercator.tiff');

  const tempPath = path.resolve(__dirname, './data/result/tiles');
  const jpegPath = path.resolve(__dirname, './data/result');
  const mbPath = path.resolve(__dirname, './data/result/mbtiles/tile.mbtiles');

  rp
    .use(
      new RasterProcess.task.ReadData(),
    )
    .use(
      new RasterProcess.task.WriteTiff(tiffPath, {
        clear: true,
        gray: false,
        bandsFunction: (info) => {
          const t = info.GRIB_PDS_TEMPLATE_ASSEMBLED_VALUES.split(' ');
          if (info.GRIB_ELEMENT === 'TMP') {
            return {
              name: 'TMP',
              label: '温度',
              process: (v) => RasterProcess.normalizeDataProcess.subScalar(v, 0),
            };
          }
          return false;
        },
        customProj4: '+proj=longlat +datum=WGS84 +no_defs +type=crs', // 4326
        customExtent: [-180, -85.05112877980659, 180, 85.05112877980659],
      }),
    )
    .use(
      new RasterProcess.task.Reproject(mercatorTiffPath, {
        width: 1024,
        height: 1024,
        clear: true,
      }),
    )
    .use(
      new RasterProcess.task.GenerateTiles(tempPath, {
        clear: true,
        gray: true,
        writeExif: true,
        zooms: [0, 1, 1],
        bandName: '',
        reprojectOptions: {
          srcNodata: NaN,
          dstNodata: NaN,
        }
      }),
    )
    .use(
      new RasterProcess.task.GenerateJPEG(jpegPath, {
        clear: true,
      }),
    )
    .use(
      new RasterProcess.task.UploadOSS({
        // region填写Bucket所在地域。以华东1（杭州）为例，Region填写为oss-cn-hangzhou。
        region: process.env.region,
        // 阿里云账号AccessKey拥有所有API的访问权限，风险很高。强烈建议您创建并使用RAM用户进行API访问或日常运维，请登录RAM控制台创建RAM用户。
        accessKeyId: process.env.accessKeyId,
        accessKeySecret: process.env.accessKeySecret,
        // 填写Bucket名称。
        bucket: process.env.bucket,
        folder: 'tmp',
        pathFunction: (url, folder) => path.join(folder, url.replace(jpegPath, '')),
      })
    )
    .run([dataPath], (err, res) => {
      console.log(res);
    });
}

run();

Task 列表

Task Name	Desc
ReadData	Read data using GDAL
WriteTiff	Write raster data to tiff
Reproject	Reproject raster data to another projection
GenerateTiles	Generate Mercator raster tiles
GenerateJPEG	Generate jpeg file
GeneratePNG	Generate png file
UploadOSS	Upload generated raster data to oss
WriteMBTile	Generate mbtile file

示例

ReadData

Options:

Option	Type	Desc
autoClose	boolean	是否自动关闭 Dataset，一般情况下我们不需要配置此参数，在链式操作中 Dataset 不需要关闭

const rp = new RasterProcess();
const res = (await rp.use(new RasterProcess.task.ReadData()).run([dataPath]));
expect(res[0]).toEqual(dataPath);

WriteTiff

Options:

Option	Type	Required	Desc
clear	boolean	false	是否自定清除已存在的 tiff 数据，默认是 false
width	number	false	默认是dataset 数据的 width
height	number	false	默认是dataset 数据的 height
dataType	enum	false	写入 Tiff 的数据类型，默认是 GDT_ Float32
bandCount	number	false	写入 Tiff 的数据的通道数，默认是原有栅格数据的通道数
bandsFunction	Function	false	是否过滤原有数据的通道，默认所有通道都会返回 true, 你可以配置此项来过滤哪些通道是你需要写入到 tiff 数据的
gray	boolean	false	是否转为灰度数据
drivers	enum	false	默认为GTiff
customProj4	string	false	自定义数据投影，默认从原始数据读取
customExtent	number[]	false	自定义数据范围，默认从元数据读取

const rp = new RasterProcess();
await rp
  .use(new RasterProcess.task.ReadData())
  .use(
    new RasterProcess.task.WriteTiff(tiffPath, {
      clear: true,
      // customProj4: '+proj=longlat +datum=WGS84 +no_defs +type=crs', // 4326
      // customExtent: [-180, -85.05112877980659, 180, 85.05112877980659],
    }),
  )
  .run([dataPath]);

Reporject

clear: boolean;
width: number;
height: number;
dataType: string;
bandCount: number;
drivers: string | string[];
resampling: string;
sourceProj4: string;
threads: number;
withMetadata: boolean;
sourceExtent: [number, number, number, number];
destinationProj4: string;
destinationExtent: [number, number, number, number];
cutline?: Geometry;
srcBands?: number[];
dstBands?: number[];
srcAlphaBand?: number;
dstAlphaBand?: number;
srcNodata?: number;
dstNodata?: number;
blend?: number;
memoryLimit?: number;
maxError?: number;
multi?: boolean;

Options:

Option	Type	Required	Desc
clear	boolean	false	是否自定清除已存在的投影后的数据，默认是 false
width	number	false	默认是256
height	number	false	默认是256
dataType	enum	false	投影后的 DataSet 数据类型，默认是 GDT_ Float32
bandCount	number	false	写入 DataSet 的数据的通道数，默认是原有栅格数据的通道数
drivers	enum	false	默认为GTiff
resampling	enum	false	重投影的重采样方式，默认是GRA_NearestNeighbor
sourceProj4	string	false	手动指定源数据投影，默认从源数据读取
sourceExtent	number[]	false	手动指定源数据范围，默认从源数据读取
withMetadata	boolean	false	是否同步元数据
destinationProj4	string	false	手动指定目标投影，默认是墨卡托 proj4定义字符串
destinationExtent	number[]	false	手动指定目标数据范围，默认根据目标投影从源数据数据范围转换
threads	number	false	重投影使用的线程数，默认使用os.cpus().length，但是需要注意，默认的配置在虚拟机或者 docker 中可能会出问题，如果出现问题建议手动配置此参数
others			其他配置项，请参照 gdal-async 文档的 reprojectImage 配置项

const rp = new RasterProcess();
await rp
  .use(new RasterProcess.task.ReadData())
  .use(
  new RasterProcess.task.Reproject(mercatorTiffPath, {
    width: 1024,
    height: 1024,
    clear: true,
    destinationProj4:
    '+proj=merc +a=6378137 +b=6378137 +lat_ts=0 +lon_0=0 +x_0=0 +y_0=0 +k=1 +units=m +nadgrids=@null +wktext +no_defs +type=crs', // 3857
    destinationExtent: [
      -20037508.342789244, -20037508.342789255, 20037508.342789244, 20037508.342789244,
    ],
  }),
)
  .run([dataPath]);

GenerateTiles

clear: boolean;
tileSize: number;
zooms: [number, number, number] | [number, number] | number;
dataType: string;
bandCount: number;
drivers: string | string[];
clipExtent: boolean;
writeExif: boolean;
gray: boolean;
bandName: string | ((zoom: string, band: number, info: any) => string);
tileFolder: string;
cacheFolder: string;
cacheFilePrefix: string;
enlargeOptions: Partial<IEnlargeDataOptions>;
tileExtent: [number, number, number, number];
tileProj4: string;
reprojectOptions: Partial<IReprojectOptions>;

Options:

Option	Type	Required	Desc
clear	boolean	false	是否自动清除已存在的瓦片数据和投影缓存数据，默认是 false
tileSize	number	false	瓦片大小默认是256
dataType	enum	false	生成瓦片的 DataSet 数据类型，默认是 GDT_ Float32
bandCount	number	false	写入 DataSet 的数据的通道数，默认是1
drivers	enum	false	默认为GTiff
gray	boolean	false	是否转灰度
writeExif	boolean	false	是否写入 exif 信息
clipExtent	boolean	false	在生成墨卡托瓦片时如果瓦片的投影范围超过墨卡托规定的投影范围自动裁剪
bandName	string、Function	false	指定栅格数据的通道名称，这在处理GFS气象数据时非常有用
tileFolder	string	false	生成瓦片的文件夹，默认是tiles
cacheFolder	string	false	在生产瓦片时我们需要针对每一级的原始数据做重投影，但是我们不需要每次都进行生成，可以在每次生成后通过文件缓存下来，这个配置就是指定了缓存的位置，当然我们也可以配置 clear=true 不进行缓存，此配置项默认是cache
cacheFilePrefix	string	false	缓存文件名称前缀，此配置项默认是mercator
tileProj4	string	false	每一层级重投影瓦片的配置，默认是墨卡托的 proj4
tileExtent	number[]	false	每一层级重投影瓦片的配置，默认是墨卡托的全球范围
reprojectOptions	object	false	每一层级重投影瓦片的重投影配置，请参照Reporject 的配置项
enlargeOptions	object	false	每一层级重投影瓦片外扩配置

const rp = new RasterProcess();
await rp
  .use(new RasterProcess.task.ReadData())
  .use(
  new RasterProcess.task.GenerateTiles(tilesPath, {
    clear: true,
    gray: true,
    writeExif: true,
    zooms: [0, 1, 1],
  }),
)
  .run([path.resolve(__dirname, './fixtures/gfs.t12z.pgrb2.0p25-write-mercator.tiff')]);

GenerateJPEG

 clear: boolean;
 quality: number;
 drivers: string | string[];
 tileFolder: string;
 name: string | ((p: string, t: string, n: string, z?: number, x?: number, y?: number) => string);

Options:

Option	Type	Required	Desc
clear	boolean	false	是否自动清除已存在的瓦片数据，默认是 false
quality	number	false	图片质量，默认是 90
drivers	enum	false	默认为JPEG
tileFolder	string	false	生成图片存放的文件夹，默认是jpeg
name	string、Function	false	生成的图片名称

const rp = new RasterProcess();
await rp
  .use(new RasterProcess.task.ReadData())
  .use(
  new RasterProcess.task.GenerateJPEG(jpegPath, {
    clear: true,
  }),
)
  .run([path.resolve(__dirname, './fixtures/gfs.t12z.pgrb2.0p25-write-mercator.tiff')]);

GeneratePNG

 clear: boolean;
 quality: number;
 drivers: string | string[];
 tileFolder: string;
 name: string | ((p: string, t: string, n: string, z?: number, x?: number, y?: number) => string);

Options:

Option	Type	Required	Desc
clear	boolean	false	是否自动清除已存在的瓦片数据，默认是 false
quality	number	false	图片质量，默认是 90
drivers	enum	false	默认为PNG
tileFolder	string	false	生成图片存放的文件夹，默认是png
name	string、Function	false	生成的图片名称

const rp = new RasterProcess();
await rp
  .use(new RasterProcess.task.ReadData())
  .use(
  new RasterProcess.task.GenerateJPEG(jpegPath, {
    clear: true,
  }),
)
  .run([path.resolve(__dirname, './fixtures/gfs.t12z.pgrb2.0p25-write-mercator.tiff')]);

UploadOSS

 region: string;
 accessKeyId: string;
 accessKeySecret: string;
 bucket: string;
 endpoint: string;
 secure: boolean;
 internal: boolean;
 cname: boolean;
 folder?: string;
 headers?: {
 [key: string]: any;
 };
 pathFunction?: (url: string, folder?: string) => string;

Options:

Option	Type	Required	Desc
region	string	true	Bucket所在地域。以华东1（杭州）为例，Region填写为oss-cn-hangzhou
accessKeyId	string	true	阿里云账号AccessKey拥有所有API的访问权限，风险很高。强烈建议您创建并使用RAM用户进行API访问或日常运维，请登录RAM控制台创建RAM用户
accessKeySecret	string	true	阿里云账号AccessKey拥有所有API的访问权限，风险很高。强烈建议您创建并使用RAM用户进行API访问或日常运维，请登录RAM控制台创建RAM用户
bucket	string	true	存储桶名称
endpoint	string	false	eg: oss-cn-hangzhou.aliyuncs.com
secure	boolean	false	是否启用 https
internal	boolean	false	是否启用内网
cname	boolean	false	默认为false，使用自定义域名访问oss。如果为true，则可以使用自定义域名填充端点字段
folder	string	false	存储目录
headers	object	false	请查看ali-oss
pathFunction	function	false	文件存储路径

const rp = new RasterProcess();
const res: any[] = await rp
  .use(new RasterProcess.task.ReadData())
  .use(
  new RasterProcess.task.UploadOSS({
    // region填写Bucket所在地域。以华东1（杭州）为例，Region填写为oss-cn-hangzhou。
    region: process.env.region as string,
    // 阿里云账号AccessKey拥有所有API的访问权限，风险很高。强烈建议您创建并使用RAM用户进行API访问或日常运维，请登录RAM控制台创建RAM用户。
    accessKeyId: process.env.accessKeyId as string,
    accessKeySecret: process.env.accessKeySecret as string,
    // 填写Bucket名称。
    bucket: process.env.bucket as string,
    folder: 'tmp',
    pathFunction: (url, f) => path.join(f as string, url.replace(jpegPath, '')),
}),
  )
    .run([path.resolve(__dirname, './fixtures/gfs.t12z.pgrb2.0p25-write-mercator.tiff')])

WriteMBTile

 region: string;
 accessKeyId: string;
 accessKeySecret: string;
 bucket: string;
 endpoint: string;
 secure: boolean;
 internal: boolean;
 cname: boolean;
 folder?: string;
 headers?: {
 [key: string]: any;
 };
 pathFunction?: (url: string, folder?: string) => string;

Options:

Option	Type	Required	Desc
mode	enum	true	Mode type

const rp = new RasterProcess();
await rp
  .use(new RasterProcess.task.ReadData())
  .use(
  new RasterProcess.task.GenerateTiles(tilesPath, {
    clear: true,
    gray: true,
    writeExif: true,
    zooms: [0, 1, 1]
  }),
)
  .use(
  new RasterProcess.task.GeneratePNG(pngPath, {
    clear: true,
  }),
)
  .use(
  new RasterProcess.task.WriteMBTile(mbPath, {
    mode: 'rwc',
  }),
)
  .run([path.resolve(__dirname, './fixtures/gfs.t12z.pgrb2.0p25-write-mercator.tiff')])