No abstract

No abstract

The Disaster Monitoring Constellation (DMC) is an international micro-satellite constellation, currently capable of providing 32m spatial-resolution, tri-band, multi-spectral imagery over a 600km swath of any location in the world, within a 24-hour period. These satellites are soon to be augmented by further spacecraft, each carrying an additional 4m resolution panchromatic imager. The spectral band selection for the DMC spacecraft has been deliberately chosen to be equivalent to that of the SPOT HRV multi-spectral imagers and bands 2, 3 and 4 of the LANDSAT Thematic Mapper. Previous satellites have used these bands in a multitude of application areas from land-use classification, crop yield assessments and vegetation stress monitoring, to burn scar mapping, urban growth modeling and insect threat assessment. However, the unique combination of high-swath, medium-spatial resolution and high-temporal resolution of the DMC opens up a new range of application areas: The system is found to be ideal for monitoring the development of fire risk over very short temporal scales using a combination of scene-derived vegetation indices, classification maps and external thermal data. Additionally, the DMC data provides a unique resource for higher resolution mapping of the fire effects (burn scars), without the need of extensive "stitching" of images. Similarly, timely data on flooding can be obtained, and changes monitored on a daily basis. Already the DMC satellites have provided image data for humanitarian relief efforts in Darfur and following the recent Asian tsunami.

Compression reduces the amount of data to be sent to ground whilst preserving its content, allowing a lower bit-rate link to be used or more data to be sent over the same link. This paper presents the results of an investigation into image compression methods for multispectral images for use on board small satellites. An error resilient compression scheme is proposed and implemented. A modification of a neural network based data compression method is discussed. The investigated compression methods are compared in terms of error resilience, compression ratio and execution time.