Their onboard computers generally perform other tasks in addition to data acquisition and manipulation.
Small satellites also require smaller and lower-cost launch vehicles. They offer benefits over larger satellites in terms of cost, development time, and payload modularity. Picosatellites such as CubeSats are excellent platforms for education as well as technology demonstration and are thus extremely valuable for countries without fully funded space programs.
The adoption of small satellites began in the year 2000 with the emerging era of “new space” as Stanford University launched its microsatellite, called the Orbiting Picosat Automated Launcher (OPAL), containing six picosatellites. These significant band acquisitions result in large three-dimensional hyperspectral images, which make their onboard compression mandatory, especially for small satellites where the platforms are confined to limited storage capacity, weight, and power budget.Ĭonceptual view of the construction of hyperspectral images by a pushbroom scanner (not shown) during airborne flights. According to, the number of bands of recognized hyperspectral imagers is as follows: (1) as many as 316 bands are acquired by the two payloads carried in the Indian Hyperspectral Imaging Satellite (HySIS) (2) 240 bands are collected by the Italian space agency’s satellite called no other than PRISMA, for PRecursore IperSpettrale della Missione Applicativa (3) 220 bands are collected by the Hyperion imager onboard NASA’s Earth Observation satellite (EO-1) and (4) 232 bands are acquired by the German mission, known as the Environmental Mapping and Analysis Program (EnMAP). The spatial dimensions ( x and z) of the hyperspectral image are constructed one scan line at a time during flight time. As illustrated in this figure, the y dimension represents the number of bands, and the x dimension corresponds to the swath width of the scene. Typically, hyperspectral images consist of hundreds of contiguous bands, and the number of these bands depends on the detector resolution (see Figure 1). The Atmospheric InfraRed Sounder (AIRS) is not far from the latter and can yield about 12 Gigabytes of data per day. For instance, the Airborne Visible InfraRed Imaging Spectrometer (AVIRIS) produces data as large as 16 Gigabytes per day. The richness of information in hyperspectral images and the enhancements in sensor performance present an ever-increasing challenge due to the large size of hyperspectral data. According to the research conducted by BCC (Business Communications Company, Wellesley, MA, USA), the growth of the global market for HSI is expected to increase at a Compound Annual Growth Rate (CAGR) of 14.7% for the period 2018–2023, from $104.0 million in 2018 to $206.2 million in 2023. It has been steadily growing over the last few years. Such applications include environmental monitoring, agricultural field observation, geological mapping, and mineral exploration, to name just a few. Hyperspectral Imaging (HSI) is an enabling technology for a variety of remote sensing applications related to intelligence, commerce, agriculture, military, and even humanitarian purposes. We conclude by highlighting some of the research gaps in the literature and recommend potential areas of future research. Furthermore, we rank the best algorithms based on efficiency and elaborate on the major factors impacting the performance of hardware-accelerated compression. We present a comparative performance analysis of the synthesized results with an emphasis on metrics like power requirement, throughput, and compression ratio.
We reviewed a total of 101 papers published from 2000 to 2021. We present herein a systematic review of hardware-accelerated compression of hyperspectral images targeting remote sensing applications. The availability of onboard data compression would help alleviate the impact of these issues while preserving the information contained in the hyperspectral image. This is particularly crucial for small satellite applications where the platform is confined to limited power, weight, and storage capacity. Issues pertaining to bandwidth limitation also arise when seeking to transfer such data from airborne satellites to ground stations for postprocessing. It requires significant processing power and large storage due to the immense size of hyperspectral data, especially in the aftermath of the recent advancements in sensor technology. Hyperspectral imaging is an indispensable technology for many remote sensing applications, yet expensive in terms of computing resources.