Assessment of lossless compression algorithms and their performance on near-infrared spectral images

Storing and streaming hyperspectral images (HSI) requires significant resources. One suitable strategy of reducing this resource utilizations is the use of compression algorithms. In this work, we benchmarked Bzip2, Gzip, LZ4 and lzf algorithms on a dataset of HSIs of lightweight plastic packaging and achieved a lossless compression as high as 96.7% compared to the original filesize.

OCM 2025 - Optical Characterization of Materials: Conference Proceedings