1996/97 Mars Sketch (9)
from CMO #206 (25 August 1998)

-- NISHITA's CCD Images of the 1997 Mars
at the Fuku City Observatory --

Japanese here

  Akinori NISHITA (Ns) tried several times from February to April 1997 to take CCD images of Mars using the 20cm refractor of the Fukui City Observatory. Here he shows an example of a raw image and its improved processed images. Figures 1, 2 and 3 show respectively the "original" image taken on 5 March 1997 at 15:00GMT (ω=299░W), its enhanced image by the unsharp masking method, and the final image processed by the Maximum Entropy Method (MEM).

NISHITA's CCD Image on 5 Mar 1997 (093░Ls) at ω=299░W, φ=25░N

  The cooled camera attached was a Mutoh CV-04 (Japanese made), and the PC brought in the dome was a portable PC in a Notebook size with a liquid crystal monitor (Toshiba TECRA720 with a CPU of Pentium 133). The filter used was only R60 (sharply cut below 600nm) and each shot was made with a 0.3 second exposure working at f/120. NISHITA made about 40 shots during a period of twenty minutes and secured a dozen of the better images, and he repeated this procedure every forty minutes.

  The exposure time was, however, short, as pointed out by Gianni QUARRA on the occasion of the CMO Meeting in January. The camera can save any image as a 16-bit file, but NISHITA's images didn't convey well the dynamic range but were like 8-bit images even at the brightest part because of the shorter exposure. The images which Ns showed in the January Meeting were just improvements over the raw images by the use of the unsharp mask.

  NISHITA then used the software Stellar Image (recommended by Tomio AKUTSU) to add several images taken at nearly the same time in order to increase the density. The "original" image thus obtained is the one shown in Fig 1. The procedure of adding images improves the S/N ratio.

  The image in Fig 2 is an improved one of Fig 1 by the use of unsharp masking of the Photoshop. The image contrast is greater, but the image is grainier.

  NISHITA and Yasunobu HIGA, in their joint private seminar on the occasion of the Meeting, then tried to use the Maximum Entropy Method to improve the images. They used the software "Maxlm DL2", which corresponded to the Hidden Image reconstructed for Windows 95. HIGA had an impression that the images processed by the "Maxlm DL2" produced a quality similar to that seen in the images presented by G QUARRA at the Meeting where QUARRA demonstrated the processing of NISHITA's raw image by the use of the MIPS software.

  The images newly presented here by NISHITA are based on the images obtained by summing multiple exposures. The MEM by "Maxlm DL2" is performed roughly as follows: Use first [Histogram] and choose [Planet]. Then setup every term in [MaxEnt]. Next choose [Select From Image] in [Set Point-Spread Function]. No use of [Automatic Select]. Just find a good point in [Point Set] to get a better image by trial and error. Use successfully [Auto-Extract] for [Set Noise] in [MaxEnt]. The background may be also fixed by setting up [Extract From Image]. Process then by [Run MaxEnt]. If unsuccessful, one should try again and repeat to use [Set Point-Spread Function] to set up differently. The [Number of Iterations] must be fixed in [Run MaxEnt]. This number must also be found through trial and error.

  The compiled 16-bit file must be converted to a 8-bit file because the printer cannot produce so much graduation. For conversion, NISHITA also used "Maxlm DL" and chose [8-bit] for [Output Range] in [Stretch] of [Process]. The [Input Range] is fixed in [Screen Stretch] or in [Manual Settings]. NISHITA saves the secured images in TIFF format with a .TIF extension. This file was found not to be read by "Photoshop v4", and so NISHITA used "Paint Shop Pro v4.2" and overwrote.

  Unsharp masking is easily done by the use of "Photoshop v4": Choose the sequence: [filter]-- [sharp]-- [unsharp mask]. Also, slightly supplement by [image]-- [adjust]-- [brightness/contrast].

  NISHITA writes here that in the case of the Moon, the surface structure can be enhanced and made sharper by the unsharp masking method. However, in the case of Mars, the MEM may be much preferred. If a bright point star is included within the frame (for example in the case of the planetary nebulae), the MEM is quite easily performed because the parameters are fixed in an easier way.

  NISHITA's trial in 1997 was, however, insufficient. He did not use any blue filter simply because no nice filter was found at that time and, furthermore, the sensibility of the CV-04 is weak in the violet region. He hopes to do better in the coming 1999 apparition based on the experience he got in 1997 (imaging) and 1998 (processing).

  Finally NISHITA expresses his sincere thanks to T AKUTSU and to Y HIGA and not least to G QUARRA for their help shown to him on the occasion of the 6th Meeting of the CMO Planetary Observers held on 2/3 January 1998.


Back to the CMO Home Page/Back to the Fašade