Project Phoenix update
- Our volunteer project programmer, David Scheidt, not a ham as of yet but hopefully soon, has written Java code that reads a data stream of 32x48 images and displays it as progressive image frames. We have modified the NBTVA standard to define the image as being framed with an all black perimiter one pixel wide and a nesting all white perimiter one pixel wide. There are no sync pulses as such. The Java software examines the data stream, finds the perimiter and displays the images.
At this time, the software can integrate the data from a still picture of a test card containing 12% random noise and extract a perfect image. His software shows three images on the computer screen: one of raw incoming data, a second of formatted data in the form of image frames, and a third of analyzed data with the intended image extracted from the noise. It's fascinating to watch this program interpret a signal of a noisy test card and see the clear picture emerge from the noise.
At present, the data is either black or white, not shades of gray, and is in the form of text files for convenience sake. The next developments will be increasing the integration levels to see how much random noise can be present and still allow discernment of the intended image, and to move from text files to graphic imagery.