using bitmaps for storing and analyzing numeric data (use case: flight sim game AI)

Question

I had idea of storing flight mission data in bitmaps for analysis.

Data stored could be velocity vector of our aircraft, velocity vector of target, location relation vector between the two, aim lead angle as a vector... and maybe a few more such as altitudes, etc. we could end up in 30 or so parameters that we want to store. At least for most the 0...255 resolution is enough. With each RGB pixel storing 3 parameters we could have snapshot of the situation stored in an 12 pixel image with room to spare. A small 64x64 image could store 340 samples (seconds) of gameplay. This would be a long enough to represent an encounter in the game. So say I then fly (with human as a pilot) gazillion missions and store these pictures as samples of each encounter.

Goal would be to have a system that would look at current "12 pixel sample" and based on that tell us what the best next "12 pixels" options are.

Are there any benefits in storing data in image format in terms of analyzing the data and suggesting next successful action. Naturally same data can be just tables or similar but would the image format give allow leveraging existing image tools or similar benefits?

cheers. And by the way I am a total amateur when it comes to algorithms etc. But would still appreciate any input.

Answer 1

The image processing and related models like CNNs are using the fact that pixels close together in an image are more related and small translations in objects are not important so using images do not necessarily improve results as there are also few pixels. Another problem may be when data is stored, it may be compressed such as by jpeg if not careful and lose data. Otherwise, should have little overhead (from image headers) compared to pure binary files. Text files on the other hand would likely be a bit larger but would be much more readable (and flexible).