RESEARCH ARTICLES CURRENT SCIENCE, VOL. 112, NO. 6, 25 MARCH 2017 1158 *For correspondence. (e-mail: kurian_mathew@sac.isro.gov.in) Correction of Mars Colour Camera images for identification of spectral classes Kurian Mathew*, A. S. Arya, Harish Seth, S. Manthira Moorthi and P. N. Babu Space Applications Center, Indian Space Research Organisation, Ahmedabad 380 015, India Mars Colour Camera on-board the Mars Orbital Mis- sion makes use of a Bayer pattern detector. Spectral response of RGB (red, green and blue) pixels of Bayer detector shows large overlap which reduces the spec- tral information content of the image. In the present paper, a simple method is suggested to correct the data for spectral overlap. It is shown that correction process significantly increases the spectral informa- tion content of the image and enhances the ability of the sensor to identify different target types like dust clouds and water ice clouds. Keywords: Bayer-pattern filters, dust clouds, ice clouds, Mars Colour Camera, spectral overlap. MARS Colour Camera (MCC) onboard the Mars Orbiter Mission (MOM) is a medium resolution imager that makes use of a Bayer pattern detector 1 . MOM was inserted into the Mars orbit on 24 September 2014 and since then MCC has acquired about 700 images. MCC images are of good radiometric and spatial quality and have been used for mapping various morphological fea- tures of Mars as well as for monitoring Martian weather phenomena like dust storms, dust devils, clouds, etc. 2–5 . They also provide context information for other science payloads on-board MOM. Though MCC consists of three spectral bands (red, green and blue), it is difficult to extract spectral information from image data since spec- tral response of RGB bands defined by Bayer filters shows large overlap. We discuss a procedure which corrects the spectral overlap and derive the radiance approximately in three non-overlapping spectral bands. Through simulation studies, we show that this procedure improves the spectral information content of the images. MOM orbits in a highly elliptical path with orbital elements evolving slowly over time. Apoareon height, periareon height and period of the orbit varied between 71,000 and 80,000 km, 250 and 500 km and 65 and 74 h respectively during the present study period. Imaging from apoareon positions enables MCC to have synoptic view of entire Mars disc and surrounding atmosphere. MCC has acquired many full disc images of Mars with its illuminated limb showing cloud and dust structures with great clarity. Formation of clouds and its dynamics in the Martian atmosphere are of great scientific interest but yet to be understood fully 6–8 . Identification of different cloud types (dust clouds, water ice clouds and CO 2 ice clouds) is important in this regard. Cloud types are usually differ- entiated from their spectral signature in the short wave in- frared region. For example, CO 2 ice can be readily identified by its characteristic absorption at 1.9 m. Even though observations in the visible wavelengths do not give conclusive answers, they may provide valuable clues about cloud composition. MCC with its broad and over- lapping spectral bands is not designed for spectral identi- fication of scenes. Yet, considering the fact that MOM mission does not have any imaging spectrometer or mul- tispectral camera in the visible wavelength range, it is important to utilize the spectral information available from MCC images. We also discuss tentative identifica- tion of dust and ice clouds from their spectral signatures as well as from other auxiliary information like its geographic location and local time of observation. MCC data corrected for spectral overlap was used for this pur- pose. Correction for spectral overlap MCC is a medium resolution camera which makes use of a 2048  2048 element Bayer-Pattern colour detector. Table 1 gives salient features of MCC. Unlike a multi- spectral camera where separate detectors and filters are used to measure radiance in different spectral bands, MCC uses a single Bayer detector thereby greatly simpli- fying camera design. In a Bayer detector RGB colour fil- ters are arranged on the square grid of photo sensors as shown in the inset of Figure 1. Since spectral separation is achieved using micro-filters fabricated over individual pixels, it is difficult to attain well-defined narrow spectral bands. The main panel of Figure 1 gives normalized spec- tral response of RGB bands of MCC 9 . As can be seen, there is lot of overlap between three bands which reduces spectral information content of images. Especially, signal measured in blue and green bands has significant contri- butions from red spectral region. For scenes which have very low reflectance in the blue region compared to red region, error in radiance measured in blue band will be large. Since almost the entire Mars surface is dark in blue