AD ALTA JOURNAL OF INTERDISCIPLINARY RESEARCH

PROCESSING OF ASTRONOMICAL IMAGES USING MATLAB IMAGE PROCESSING TOOLBOX

ELIŠKA ANNA KUBIČKOVÁ shown in Figure 2. Edge detector is usually used as a pre-
processing method before Hough transformation. We can see an
University of West Bohemia, Univerzitní 22, 30614 Pilsen original image (a), edge image (b), parametric space (c), and
email: eliskaak@kky.zcu.cz detected lines (d).

The work has been supported by the grant of The University of West Bohemia:
“Intelligent methods of machine vision and understanding”, Project No. SGS-2010-
054.

Abstract: The paper deals with searching for meteors in digital astronomical images.
Hough transformation for searching for straight lines was used. Four functions were
created on the basis of this method for various types of meteoric snaps. Graphic User
Interface (GUI) for handling of functions and the database of snaps was created on the
basis of MATLAB. MATLAB tool Image Processing Toolbox was used for image
processing as well.

Keywords: meteor detection, Hough transformation, graphic user interface, MATLAB,

Image Processing Toolbox

1 Introduction Into Processing of Astronomical Snaps

Astronomical records can be divided into some different groups.

We can distinguish static and dynamic records, classical
analogue or digital records, professional and amateur
astronomical snaps, and of course, astronomical records can be
divided by their contents. This work is aimed at processing and
analysis of static digital meteoric snaps. The task was to search
for meteors in meteoric snaps.

1.1 Hough Transformation

Meteors in astronomical snaps usually have a typical straight- Figure 2 Main steps of Hough transformation by [1]
line shape. Hough transformation is the most current method for
searching for straight lines in digital images. The principle of 1.2 Functions for Meteor Searching
Hough transformation for straight lines detection will be
explained using the straight line with slope-intercept form. A Four functions were created to search for meteors in meteoric
straight line is defined by two points: A = (x1, y1), B = (x2, y2), snaps. It was basic meteoric function, function using median
see Figure 1. Slope-intercept forms for straight lines going smoothing, function with image rotation, and combined filtration
through the points A and B are given by the equations: and rotation function. The MATLAB tool Image Processing
y1 = kx1 + q (1), Toolbox [3] contains many functions to process digital images.
y2 = kx2 + q (2), Image Processing Toolbox functions were used to create new
when k, q are parameters. These equations are interpreted in the functions for meteor searching. Detailed description of Image
parametric space for parameters k, q: Processing Toolbox functions and new created meteor functions
q = -kx1 + y1 (3), was brought in the paper [2]. We can briefly summarize that
q = -kx2 + y2 (4). main Image Processing Toolbox functions used in meteor
The straight lines given by the equations (1) and (2) in the searching are the followings: greyscale transformations,
Cartesian coordinated system are given by the equations (3) and geometric transformations, image smoothing (especially median
(4) in the parametric space. The only common point of both filtering), and edge detector. These functions are further briefly
straight lines in the parametric space is the point, which described.
represents the only existing straight line connecting points A and Greyscale transformations are processes, which do not depend
B in the original image space. It is the main principle of Hough on the position of the pixel in the image. Greyscale
transformation for searching for straight lines that all straight transformation functions transform the original image with the
lines and their parts in the image space are transformed into the given brightness into an image with a new brightness. Geometric
only points in the parametric space. transformations realize basic geometric operations with the
image: rotation, change of the scale and skewing by the angle.
Image smoothing is an image pre-processing method, which is
used to suppressing of noise in the image. Image smoothing
methods, which are not edge preserving, blur sharp edges, which
causes lost of information. Median filtering is a non-linear
smoothing method that reduces the blurring of edges. The
current image point is replaced by the median of the brightnesses
in its neighbourhood. Edge detector is a general tool for finding
of lines in the image. MATLAB uses edge detector, which is
based on the local convolution in the image with convolution
Figure 1 Principle of Hough transformation kernels, which serve as line patterns. It is used as a line finding
operator. All described functions were used in new created
All image points are transformed into some points in the functions. The particular realizations of these functions are
parametric space. All values in the parametric space are implemented by the user menu for meteor searching, which is
quantized and the parametric space is divided into elementary described in the chapter 2.
cells. Points, which belong to the elements of the straight lines,
are accumulated in these cells. At the end of this process the 2 Graphic User Interface for Meteor Searching
contents of all cells are evaluated. If some cell contains
numerous points, it is a big probability that these points lie on The process of meteor searching itself is realized by the Graphic
the same line. Main steps of the Hough transformation are User Interface (GUI), which was created on the basis of the
AD ALTA JOURNAL OF INTERDISCIPLINARY RESEARCH

MATLAB GUI. This tool enables to build an interactive user contain very bright meteors. The main problems, which arose
menu, in which it is possible to insert functions, data, and during searching for meteors, are the followings: feeble meteor
databases. It is possible to separate two main steps in the user on the background with bright stars, meteor on the day sky with
menu building. First step is the choice of buttons from starting the Sun, false lines in the image, earthly objects as buildings,
menu; see Figure 3. Second step is programming of user trees, cars, mountains, various pylons. Satellites on the night sky
functions and creation of the linkage between these functions. represent a big problem as well. Special case is a very bright
Property Inspector serves for setting of technical parameters as a meteor named bolid, or fireball, which often has not a typical
size, fonts, colours, position, etc. Programming of the menu straight-line shape. These meteors could not be to search for by
functions, which belong to the buttons, sliders, and other menu Hough transformation. A few snaps with meteor showers were
items, is realized in M-file Editor (Figure 3). Resulting menu, processed as well, but processing of snaps with meteor showers
which was built to search for meteors in meteoric snaps, is in was not the goal of this work. It is a more complicated problem
Figure 4. This menu contains push buttons to call meteor demanding different methods. The most common types of
functions, pop-up menus to look through the databases of meteors were successfully found using combination of basic
processed images, list box to choose meteoric snap to process, meteor functions with median filtering or image rotation. The
and list box to look over results of the searching. The push typical meteoric snap and result of the search for meteor is
button for insertion of a new image to process is placed in the shown in Figure 5.
menu as well. It is possible to monitor the whole process of
searching for every single meteor, because partial results of
meteor detection process are available.

Figure 5 Example of detected meteor

4 Conclusions

Two hundred meteoric snaps were processed using described

functions and created menu. Almost 80% of meteors were
successfully found. The main problems mentioned above require
new methods of searching for meteors. Improving of the user
menu is necessary as well. The next step in meteor searching,
after solution of current problems, is to automatize the meteor
searching process.

Literature:

1. Sonka, M., Hlavac, V., Boyle, R.: Image Processing, Analysis,

and Machine Vision. Third Edition, Stamford: Cengage
Learning, 2008. International Student Edition: ISBN-13: 978-0-
495-24438-7, ISBN-10: 0-495-24438-4.
2. Kubičková, E.A.: Searching of Meteors in Astronomical
Images on the Basis of MATLAB Toolbox. Technical Computing
Bratislava 2010, 18th Annual Conference Proceedings,
Bratislava: RT Systems in cooperation with Systémy
Figure 3 Building of the menu on the basis of MATLAB GUI priemyselnej informatiky, s.r.o., 2010. ISBN 978-80-970519-0-
7.
3.Image Processing Toolbox, User’s Guide. The MathWorks,
Inc., 2008. www.mathworks.com.

Primary Paper Section: I

Secondary Paper Section: BN

Figure 4 Menu for meteor searching

3 Results of Meteor Searching

This chapter describes some typical results of meteor searching.

All used functions are realized on the basis of Hough
transformation. The basic meteoric function is based only on this
method. Very bright meteors were successfully found using the
basic meteor function. But numerous meteoric images do not