lerc

What is lerc?

The 'lerc' npm package is used for encoding and decoding LERC (Limited Error Raster Compression) files. LERC is a format developed by Esri for compressing raster data with controlled loss, making it suitable for geospatial applications where precision is important.

What are lerc's main functionalities?

Decode LERC data

This feature allows you to decode LERC data from a file. The decoded data can then be used for further processing or analysis.

const lerc = require('lerc');
const fs = require('fs');

// Read LERC file
const lercData = fs.readFileSync('path/to/lerc/file.lerc');

// Decode LERC data
const decodedData = lerc.decode(lercData);
console.log(decodedData);

Encode data to LERC format

This feature allows you to encode raster data into the LERC format. The encoded data can then be saved to a file for storage or transmission.

const lerc = require('lerc');
const fs = require('fs');

// Example data to encode
const data = {
  width: 256,
  height: 256,
  pixels: new Float32Array(256 * 256)
};

// Encode data to LERC format
const encodedData = lerc.encode(data);

// Write encoded data to file
fs.writeFileSync('path/to/output/file.lerc', encodedData);

Other packages similar to lerc

geotiff

The 'geotiff' package is used for reading and writing GeoTIFF files, which are a common format for geospatial raster data. Unlike LERC, GeoTIFF does not focus on compression but rather on storing georeferenced raster data. It is more versatile in terms of metadata and geospatial information but may result in larger file sizes.

pngjs

The 'pngjs' package is used for encoding and decoding PNG images. While PNG is a lossless compression format, it is not specifically designed for geospatial data. It is more general-purpose and widely supported but may not offer the same level of control over compression errors as LERC.

jpeg-js

The 'jpeg-js' package is used for encoding and decoding JPEG images. JPEG is a lossy compression format that is widely used for photographic images. It is not suitable for applications requiring precise control over compression errors, making it less ideal for geospatial data compared to LERC.

README

Lerc JS

Rapid decoding of Lerc compressed raster data for any standard pixel type, not just rgb or byte.

Breaking changes

• Web Assembly support is now required.
• Lerc.load() must be invoked and the returned promise must be resolved prior to Lerc.decode. This only needs to be done once per worker (or the main thread). There's no extra cost when invoked multiple times as the internal wasm loading promise is cached.

Browser

<script type="text/javascript" src="https://unpkg.com/lerc"></script>

await Lerc.load();

const pixelBlock = Lerc.decode(arrayBuffer);

// use options
const pixelBlock = Lerc.decode(arrayBuffer, {
  inputOffset: 10, // start from the 10th byte (default is 0)
  returnInterleaved: true // only applicable to n-depth lerc2 blobs (default is false)
});

Node

npm install lerc && npm install node-fetch-commonjs

const fetch = require('node-fetch-commonjs');
const Lerc = require('lerc');

fetch('http://elevation3d.arcgis.com/arcgis/rest/services/WorldElevation3D/Terrain3D/ImageServer/tile/0/0/0')
  .then(response => response.arrayBuffer())
  .then(async (body) => {
    // on demand load here (internally it's only loading once)
    const image = await Lerc.load().then(() => Lerc.decode(body));
    console.log(image.width); // 257
  });

API Reference

Lerc

A module for decoding LERC blobs.

load([options]) ⇒ Promise ⏏

Load the depencies (web assembly). Check whether dependencies has been loaded using Lerc.isLoaded. The loading promise is cached so it can be invoked multiple times if needed.

Kind: Exported function

Param	Type	Description
[options.locateFile]	(wasmFileName?: string, scriptDir?: string) => string	The function to locate lerc-wasm.wasm. Used when the web assembly file is moved to a different location.

decode(input, [options]) ⇒ Object ⏏

A function for decoding both LERC1 and LERC2 byte streams capable of handling multiband pixel blocks for various pixel types.

Kind: Exported function

Param	Type	Description
input	ArrayBuffer	The LERC input byte stream
[options]	object	The decoding options below are optional.
[options.inputOffset]	number	The number of bytes to skip in the input byte stream. A valid Lerc file is expected at that position.
[options.noDataValue]	number	It is recommended to use the returned mask instead of setting this value.
(Deprecated) [options.returnPixelInterleavedDims]	boolean	will be removed in next release, use returnInterleaved instead.
[options.returnInterleaved]	boolean	(ndepth LERC2 only) If true, returned depth values are pixel-interleaved.

Result Object Properties

Name	Type	Description
width	number	Width of decoded image.
height	number	Height of decoded image.
pixels	array	[band1, band2, …] Each band is a typed array of width * height * depthCount.
pixelType	string	The type of pixels represented in the output.
mask	mask	Typed array with a size of width*height, or null if all pixels are valid.
statistics	array	[statistics_band1, statistics_band2, …] Each element is a statistics object representing min and max values
(Deprecated) dimCount	number	Will be removed in next release, use depthCount instead.
depthCount	number	Depth count
[bandMasks]	array	[band1_mask, band2_mask, …] Each band is a Uint8Array of width * height * depthCount.

---

Licensing

Copyright © 2017-2022 Esri

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

A local copy of the license and additional notices are located with the source distribution at:

http://github.com/Esri/lerc/

Changelog

4.0.0 - 2022-07-15

Milestones reached

• Incremented Lerc version to 4.0 as the new, wasm based JavaScript decoder may not work on older browsers such as IE.
• Incremented Lerc version to 3.1. All existing 3.0 functions work same as before.
• Incremented Lerc codec version to v2.6.
• This LERC API and all language decoders (C++, C, C#, Python, JavaScript) and encoders (C++, C, Python) are now in sync with ESRI ArcGIS Pro 3.0.

Added

• New, additional LERC API functions *_4D(). Main reason for this upgrade is to support special cases of void or invalid data not supported yet. For 3D or 4D data with an array of values per pixel, and some of such values valid and others invalid, the new functions allow to pass one noData value per band to represent such "mixed cases". There is no change to the valid / invalid byte masks that mark each pixel as valid or invalid. There can be one mask per band, one mask for all bands, or no mask. For more details see the Readme.md, Lerc_c_api.h, and Lerc_ByteStream_Specification.pdf, in increasing level of detail. Main.cpp has some sample calls.

• New lossless compression for data types float and double.

• Support for large integers > 32 bit up to 53 bit. Pass them as double.

• Renamed nDim to nDepth, the size of the array per pixel.

• New #define ENCODE_VERIFY in Defines.h. If enabled, Lerc, as an added last step in encode, will decode the compressed blob again and compare it against the input data.

• Using Emscripten, we compiled the Lerc C++ code into web assembly, resulting in a new JS Lerc decoder. From now on, updates to the Lerc C++ code will be converted to JS automatically.

• Added a new function to the LERC API called 'lerc_getDataRanges(...)'. It allows fast access to the data ranges in a compressed Lerc blob without having to decode it. It returns 2 double arrays with the minimum and maximum values per band and depth.

Fixed

• The older Lerc codec v2.2 (used around 2016) required 3 extra padding bytes at the end of the Lerc blob. If those padding bytes are missing, Lerc decode could fail. With this fix, decode works on such v2.2 Lerc blobs with missing padding bytes.