This section lists up the steps of image processing for this system and
considers problems at each step.
After taking a new image, a user inputs the image itself and
exposed time, approximate position in the celestial globe, approximate
area width as additional information to the system in order to check
the image.
The system receives the image and additional information and goes into
process as follows.
- Select the input image
When receiving the user's image, the system has to check if it is
really an astronomical image.
However, this step is not so important, especially for an experimental
system.
- Check and correct the quality
Then the system should grasp various errors of the image such as
a guide error, out of focus, bending, etc., and it should also correct
them for the latter matching step with star catalogs.
However, this step is not so important, either.
- Detect stars automatically
This is indispensable. Detecting bright stars is not so difficult.
However new objects to be detected should be mostly faint. Then the
system has to distinguish them from noises.
- Identify each detected star with stars in catalogs
The next step is matching between detected stars and those in catalogs.
This step will take enormous time without elaboration.
- Announce known objects in the image
The data of known objects are updated day by day. So the system uses
the latest data via network.
- Announce detected candidates of unknown objects
Finally the system tells the user the detected candidates of new
objects, if they are not in catalogs or much different from those in the
old images.
- Register the image to the database
After these steps, the system register the image to the database.
The image will be browsed by other users in the future.
|