Image-processing techniques in precisely measuring positions of Jupiter and its Galilean satellites

Abstract

Abstract. Two image-processing techniques are developed to measure precisely the positions of Jupiter and its Galilean satellites. One is used to detect the edge of Jupiter's CCD image and fit it with an ellipse, thereby obtaining the planetary center. The other is used to remove the positional effect of Jupiter's scattering light that has on the measurement of a close satellite. 133 frames of CCD images are measured by using these techniques, and these measured positions of Galilean satellites are compared to the ones computed with the ephemerides of Jacobson's JUP204 and Arlot's G5. Preliminary analysis of the data as Jupiter-satellite or inter-satellite positions shows that a standard error (i.e. internal precision) as small as 0.03 arcsec can be reached in right ascension and declination. It is believed that these techniques would be useful for astrometric and photometric CCD observations of Galilean satellites, especially for CCD observations in the forthcoming mutual phenomena of Galilean satellites.

Jovian satellite positions from Hubble Space Telescope images

“…The estimated astrometric precision of 30 mas in the paper by Peng et al (2003) obtained with ground based galilean satellite data is very close the 26 mas measured for our HST data. The estimates of precision given by Poulet and Sicardy (2001) are 70 mas for the PC chip and 120 mas for the WF chip.…”

“…In a very recent paper by Peng et al (2003), they describe ground based galilean satellite astrometry using limbfitting for Jupiter. Their method involves isolating the disk of Jupiter by creating a sub-image, choosing a threshold brightness and converting the sub-image to a binary representation, detecting the edge of Jupiter from the binary information, and fitting the edge with an ellipse by least squares.…”

CCD Positions of Saturn and Its Major Satellites From 2002-2006

“…Step 2. The above subimage was changed into a binary image with the threshold, and a four-neighborhood mask (see Appendix A in Peng et al 2003) was further used to detect all edges of Saturn and its rings. Definitely, a binary image has only two gray levels in which one gray level 0 is designed for those original gray levels less than the threshold and another gray level 1 (or some other gray level) is designed for the other gray levels.…”

数字化天顶望远镜观测图像及数据处理

“…CCD 图像量度坐标, 即质心定位. 对于星像质心定 位已经发展了许多经典的算法 [28] : 高斯拟合法 [29] 、 矩 方法(包括修正矩方法 [27,30,31] )、中值法和寻导法 [32,33] 等. 本文采用修正矩方法, 按每个像素点的亮度值 g i 做线性加权平均, 得到各星像质心的 CCD 图像量度 坐标(x, y) [27,29] .…”