Plate Object

Plate Object

Remarks

Please read through this page completely! You'll learn invaluable things.

A Plate object represents an astronomical image and its accociated descriptive information. FITS format images may be attached to a created Plate object, and the associated information in the FITS header records will be automatically transferred into the properties of the Plate. Note that many of the "standard" FITS header fields have corresponding properties within the Plate object. Most may be modified while the image is attached, and then the updated image properties may be written back out to the FITS header.

The fundamental purpose for the Plate object is to automatically determine the true celestial coordinates of the center of the image and to provide astrometric positions of the objects in the image. The resulting resolved image coordinates may be used to correct a telescope's pointing coordinates, or written into the FITS image file as World Coordinate System (WCS) data. A FITS image containing accurate WCS data may be used (carefully!) for direct astrometric measurements since the image scale, projection geometry, and center-point coordinates are accurately known. This information suffices as an astrometric plate solution. If the original FITS image has WCS data, it will be used as a starting point for PinPoint's plate solving process.

Plate Solving Process

The plate solving process is done in three steps (which can be executed in just one method call, if desired):

1. Scan the image and detect stars, generating a list of detected stars and their image coordinates in pixels. This is the Image Star List.
2. Given a reasonable approximation of the image center-point Right Ascension and Declination, retrieve a list of stars from a reference catalog. This is the Catalog Star List.
3. Given reasonably accurate angular scale factors for the image, attempt to match image stars and catalog stars. If "enough" are matched, the correct center-point celestial coordinates are known, as well as the roll angle of the image and the precise plate scale factors. Note that the signs of the plate scale factors indicate image orientation such as mirroring and flipping.

Prescription for Success

• The center-point Right Ascension and Declination must be known to within half of the field-of-view. (see All-Sky Solving below)
• The horizontal and vertical plate scale factors (arc-seconds per pixel) must be known to within about 30% of their true values. (see All-Sky Solving below)
• The image must be "reasonably" clear, with at least a few dozen stars 4 sigma above the noise.

You don't need to know the roll-angle of the camera, nor do you need to take into account mirroring or flipping of the images. The matching algorithm in PinPoint is completely insensitive to these characteristics, and will produce the correct WCS data in all cases. If you don't know the first two, use All-Sky solving (see below).

All-Sky Plate Solving (Internet required)

If you don't know the place in the sky at which an image was acquired, PinPoint's all-sky solving feature can almost always figure this out from scratch. It uses Astrometry.net's online all-sky solving service. This service has a specific Application Programming Interface (API) designed for programs to use it remotely (as opposed to people who submit images via its web pages). PinPoint uses this API to get rough estimates of centerpoint RA and Dec, as well as the plate scale (arcsec per pixel) for the image. If this is successful, the Plate.Solve() method is automatically called using the new center coordinates and plate scale. This almost always will result in a solution, and it will have a the full accuracy of PinPoint including the distortion mapping and the resulting corner to corner astrometric accuracy. For details see the Remarks section of Plate.SolveAllSky().

Magnitude Calculations - Color Band

Matched image stars whose flux is near or below the background noise are not used in the magnitude fit. Catalog magnitudes used are of the UVBRI color band for the image, set via the Plate.ColorBand property, or the red approximation if not set or set to ppUnfiltered (same thing).

• The magnitude values produced by PinPoint are estimates produced by simply fitting the observed flux of objects in the image with the published magnitude for the given color band in the reference catalog. This is not science-grade photometry! Reference catalogs generally do not contain "good" photometry (UCAC4 and Tycho-2 are exceptions). Weaker stars in the image will have statistically large errors due to the inevitable Poisson noise. If you expect good photometry on an asteroid detection near or below the noise, it is guaranteed that you will be disappointed.
• The magnitude values produced for 1-shot color images are not on any recognized color system! They are created artifically, and are suitable only as a tool for estimating limiting magnitude of plates, and extremely rough magnitudes of objects.
• If the Plate.ColorBand property is set to ppUnknown, then internally the "red" magnitude (or the closest thing to it) from the select reference catalog will be used to form a fit and calculate the Plate.Magnitude property property.

Prescription for Success

• Use only filtered images, taken in a UBVRI standard band.
• Specify the color band of the iamge via the Plate.ColorBand property.
• Use a catalog that has photometry in the same band as the image. If the catalog does not have data in the plate's color band, magnitudes will not be available.
• Do a Gaussian convolution on the image before plate solving, using the Plate.ConvolveGaussian() method. Adjust the FWHM of the Gaussian to match the FWHM of the objects in your image. This will also improve your astrometry a bit.

Supported Reference Catalogs

Synthesized Catalog Color Bands