Home Page       Wed Oct 15 02:17:40 JST 1997

Overview of PIXY system

The PIXY( PIXII) system

Practical Image eXamination and Inner-objects Identification system

aims to set up that anyone can examine their images and check whether new objects are in them, which is one of the purposes of the MISAO project. Definitely, with the image and some additional information, the system compares it with a star catalog and replies the data of every star in the image, a list of known objects like comets and candidates of new objects which are not in the catalog.

The architecture of the PIXY system is as Figure 1 shows.

  
Figure 1: System architecture

The PIXY should make it easy for anyone to examine their own images. Therefore, the system is to be implemented in Java and available via WWW. The user accesses with an applet in a specific WWW page or a special software. First of all, the user has to input the image itself and some additional information: date and time, approximate direction, and approximate width of the image (and approximate rotating angle if required).

At first, the system detects stars from the image and makes a list of x,y coordinates and brightness of each star. At the same time, it extracts stars in the neighborhood from a catalog depending on the input approximate direction and makes another list of the coordinates and magnitude of each. Then it tries to match these two lists. Here the word matching is defined as to obtain a map function which is a coordinates converter to determine the position x,y in the image is where in the chart created from a catalog. Definitely, after rotating by an angle and magnifying by rate r, the image should correspond to anywhere in the chart. The difference between the two coordinates is . The map function consists of the four parameters .

After the map function is obtained, the system calculates all known objects which should be in the image and shows the user which stars in the image are the known objects. It also shows the candidates of new objects which are not in the catalog, and remarkable stars which seem very different from the data in the catalog. Furthermore, the user can search and view other images of the same area in the database.

I have already implemented those functions as star detection from the image, matching with a catalog, and obtaining the map function. For the present, I am experimenting with the prototype system. The prototype system re-maps the star data in a catalog with the map function and makes a chart of the same area as the original image. The original image and the chart are displayed together and I can easily confirm if the system obtained a right map function.

After this, I show you the technical details and some results of experimentations on the two points: star detection and matching.