**5. Conclusions**

In this manuscript, a new transform-based algorithm for performing lossy hyperspectral images compression, named *Lossy Compression Algorithm for Hyperspectral image systems* (HyperLCA), has been proposed. The main goal of this compressor is to provide a good compression performance at a reasonable computational burden, especially for very high compression ratios that are hardly achievable by lossless compression approaches. The proposed HyperLCA algorithm has different advantages. First of all, it is able to achieve high compression ratios while preserving the most different elements of the data set, which are crucial for many hyperspectral images applications such as anomaly detection, target detection or classification. Secondly, the compression ratio to be achieved can be perfectly fixed in advance. Additionally, some extra stopping conditions, based on quality metrics, can be added in order to stop the compression if the desire minimal quality is achieved at higher compression ratios than the specified. The possibility of adding these kinds of stopping conditions also enables a progressive decoding of the compressed bitstream. Furthermore, the HyperLCA algorithm is able to independently compress blocks of pixels of the image, increasing its error resilience and making it specially suitable for applications that use pushbroom or whiskbroom sensors. Finally, the HyperLCA also has many computational advantages, including low mathematical complexity and a high level of parallelism, which differentiate it from other state-of-the-art transform-based compression approaches and make it a more viable option for applications under tight latency constraints or applications with limited computational resources, such as hyperspectral compression on-board satellites.

An extensive amount of experiments have been performed in order to evaluate the goodness of the proposed HyperLCA compressor using different calibrated and uncalibrated hyperspectral images from the AVIRIS and Hyperion sensors. The results provided by the proposed HyperLCA compressor have been evaluated and compared against those produced by some of the most relevant state-of-the-art compression solutions. Additionally, the effect produced by compressing-decompressing the image using the HyperLCA transform in anomaly detection, hyperspectral imaging classification and spectral unmixing applications has also been evaluated. All the obtained results allow to conclude that the proposed HyperLCA compressor represents a very suitable option for lossy compressing hyperspectral images, especially when it is important to achieve high compression ratios while at the same time preserving the most different elements of the data set, and when it is important to perform the compression under tight latency and computational constraints.

**Acknowledgments:** This research is partially funded by the European Commission through the ECSEL Joint Undertaking (ENABLE-S3 project, No. 692455) and the Ministry of Economy and Competitiveness (MINECO) of the Spanish Government (ENABLE-S3 project, No. PCIN-2015-225 and REBECCA (Resilient EmBedded Electronic systems for Controlling Cities under Atypical situations) project, No. TEC2014-58036-C4-4-R.

**Author Contributions:** Raúl Guerra, Yubal Barrios, María Díaz, Lucana Santos, Sebastián López and Roberto Sarmiento were all involved in the study, design and writing of this paper. Raúl Guerra and María Díaz carried out the development of the HyperLCA compressor. Lucana Santos provided reference software and knowledge about the state-of-the-art solutions for on-board compression. Yubal Barrios performed the experiments. Sebastián López and Roberto Sarmiento supervised the work and revised the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
