The blob menu
This menu only appears on image windows which have been displayed by clicking on or near a detected object (ie, blob) in another frame. "Blob" is the term used in GRIP for a connected group of pixels that has been detected in some way. Some other systems use the term ROI (region of interest) for the same thing. One aim of the present menu is to identify blobs as stars and then to be able to measure the magnitudes of some of the stars by using others as known references. There may well be non-astronomical uses too, along similar lines. More details of the procedure for magnitude estimation are available here.
The (generally small) window that appears when a detected blob is clicked on (or near) in a parent frame appears as shown above. The blob in this case is a rather faint star (Nova Aql 2009). It is magnified so that its individual pixels can be seen, each as a 6 x 6 square of pixels. The two cross hairs run through the measured centre of the blob and the heavier green dots show the boundary of the blob as initially detected. Generally the boundary does not cover the whole blob because it had to be detected at a level higher than the background. Outside the blob are the outlines of two circles, shown by less heavy green dots. These measuring circles can be used for determining the brightness of the blob accurately, by means of options on this menu.
- displays a message box containing the area and brightness of the detected object (blob). If other menu options have been used, the values entered or measured will also be displayed.
- is for making the two outer circles larger or smaller. As initially drawn the inner circle has twice the detected diameter of the blob (actually determined as the greater of the width and height of the boundary) and the outer circle has 3 times that diameter. The adjustment here sets a multiplying factor for those two diameters. The factor can range from the initial 1 up to 10 but it is unlikely that more than 3 or 4 would be needed. The important thing is that the inner cirle must completely enclose all pixels in the blob (star) which are brighter than the background. If possible the outer ring, between the two circles must contain nothing brighter than the background (such as another star). If that cannot be arranged then the brightness measurement will not be accurate and the star will not be suitable as a reference for estimating magnitudes.
- can be used once the measuring circles have been properly set, as just described. All complete pixels in the outer ring, including those on its outer boundary but not on the inner one, are measured and statistics are calculated, particularly the mean and standard deviation. This is done separately for each colour channel of the image. Then all pixels in the inner disc, including those on its boundary, are measured and the mean background from the outer ring is subtracted from each. These inner pixel values are then summed to get the true brightness of the blob (star), in arbitrary units. Again, this is done separately for each colour channel.
- prompts for entering a string of text as a label for the detected object. This is unnecessary if another option, paste data, is used.
- prompts for entering a number which is taken to be the stellar magnitude of the detected object. Reference star magnitudes can be set in this way but the next option offers a better method in most cases.
- is available if a GRIP-generated star chart has been created and a star has been clicked in that chart. It is as if the star chart has copied the star onto a clipboard internal to GRIP. Therefore the conventional keyboard short-cut for pasting also works, without having to invoke the menu first. The star details include position, identifiers (possibly several, from different catalogues) and magnitudes (possibly in several wavebands). Pasting causes a display of all the details, as if the first option in this menu had been selected.
- is a quick way of removing all star details. This might be used if the wrong star had been associated with the blob or if you no longer want to use a certain star as a magnitude reference. Manually entered identifier and magnitude will also be cleared.
For monochrome images the dialogue is slightly different, just setting one channel.
displays a small dialogue, as shown on the right. If the specified waveband is available in the star magnitude data it will be used when estimating magnitudes for the corresponding image channel, otherwise V-band magnitudes will be used.
Estimation and graphing works for monochrome (1-channel) images and RGB (3-channel) images.
is available only when an identifier has been entered for this blob (star) but not any magnitude value. All of the detected objects in the image from which the current one came will be examined. Those which have both an identifier and a magnitude value will be assumed to be standard reference stars. Those which have an identifier but no magnitude value will have their magnitudes estimated from the standard reference stars by means of least squares fitting of a straight line on a graph of logarithms of the measured brightnesses against magnitudes. GRIP measures brightness for each channel in a coloured image as well as the overall brightness calculated as the square root of the sum of the squares of the channel brightnesses. The magnitude estimates are therefore made independently for each colour channel as well as for the overall brightness. Estimation uses either the accurately measured brightnesses or the initially detected ones. The criterion for using accurate brightnesses is that all the stars with unknown magnitudes (ie, the ones to be estimated, of which there is typically only one) have been accurately measured and that there is more than one reference star that has been measured accurately. Here is an example of a result:
(The table has been cut and squashed horizontally to fit this page.)
pops up a table of measurements (id, area, brightnesses, magnitude) for all detected blobs in the original image which have an identifier, either manually entered or pasted as star data from a star chart. Area and brightness columns refer to the blob as initially detected, with a boundary that is inevitably too small. Columns on the right give the accurate brightnesses determined by using the measuring circles completely surrounding the blob.
- saves the data for all blobs/stars in the parent image from which the current blob was displayed. The file is XML, so it is easily readable as text and can easily be converted into other formats. It includes the name of the parent image file so that both the image and all the blob/star data can be reopened from the main file menu of GRIP.
- here. displays the present page if you have installed files as described