ImageMagick Picture Quality
The quality of a received Slow Scan TV (SSTV) picture may be determined by comparing, pixel-by-pixel, for identical matches in RGB color space for the customary
image displayed on an SSTV cam in JPG format. If a ham received an image perfectly, there would be a 100% match. However, with analog transmissions, this
will never happen. Considerable image degradation occurs as soon as the software, eg, MMSSTV, readies the image for transmission and losses are magnified
as the signal encounters path loss when it moves through the ionosphere.
ImageMagick is powerful software package that can compare two images pixel-by-pixel, e.g., a 320x256 pixel JPG image from an SSTV transmission. This is a total of 81,920 pixels in many SSTV pictures. The ratio
of matches to the total is expressed as a percentage. This may be viewed as a measure of picture quality received.
In practice there is significant deteriation. A 'fuzz' adjustment is available in ImageMagick. Fuzz is used to match colors which are close to the target
colors in RGB space. Colors within this distance are considered equal. Fuzz may be expressed in absolute intensity units, or as we will do for SSTV image
comparisons, expressed as a percentage of maximum possible intensity value of each pixel.
From experimentation, 10% might be a useful fuzz tolerance level for expressing the quality of a received image.
- Required upload denoted by *
- Image uploads and processing may take a minute to complete.
- Image sizes must be indentical in size, e.g., 320x256 pixels
- Image upload formats supported: gif, jpg, png, bmp
- ImageMagick used for comparison
- Demonstration
- This SSTV image was transmitted from Spain and received in Wisconsin on 14.230 MHz. A pixel-by-pixel comparison of image quality was calculated with ImageMagick. A 10% fuzz factor was applied, meaning that any pixels between the images were deemed to match if the RGB color space of the degraded image was within 10% of the original.
- The calculated result is: 63,069 pixels FAIL to match out of 81,920 total pixels, using a fuzz of 10%. The degraded image quality is 23.01%.
- Right click on each image and select the option to save the images to your desktop.
- Select each image from your desktop as an upload for this form.
- Click the "Upload images for comparison" button.
Ham Hunting
I developed this web functionality after a QSO with Dutch KYØR in Pueblo, CO. Soon after our QSO a rogue operator came on frequency spouting vile profanities and failing to ID: all illegal operations.
That gave us the idea of triangulating to generally locate the signal. I had a general idea, but on the spot, we did not succeed in finding his general area of transmission.
This section is designed to rapidly gather the necessary information to suggest an approximate lat/long for the violator, and to display visually on a map.
This preliminary information would be helpful to an Official Observer (OO) in that region who might with the help of more local hams start to pinpoint the offender, and ultimately involve the FCC in discovery.
Step #1 is to determine the lat/long of two ham radio stations which will serve as initial observers. Simply insert the call sign. You will the obtain the signed decimal degrees, where negative indicates west/south. My QTH is inserted in the first row of the subsequent table.
Step #2 is to insert the lat/long for station B, and then each station should find the best beam heading to indicate maximum strength of the offending signal and insert in the table (180 is defaulted).
Now for Step #3, copy the resulting longitude, latititude Intersection Point data into the Latitude & Longitude boxes below, and then click on "Search."
Be sure to separate the degrees, minutes and seconds by spaces, and lead each appropriately with an "N" or "W" as in the example with my QTH inserted.
Step #4: Once the general area of the country is found, you may wish to look for local hams who might be able to help take the investigation a step further. You can do so via Reverse Call Sign Look-up
from the US FCC data base.
Technical notes: The geometric formulas are calculated on the basis of a spherical earth, ignoring ellipsoidal effects. This is accurate enough for first approximation of a distant ham station. The earth is very slightly ellipsoidal. Using a spherical model gives errors typically up to 0.3%.