Once you have the image in a viewable format (tiff, jpg or whatever) there are free stitching tools. Autostitch and Hugin are two. With these you can generate multi-megapixel images Not quite so easy, you can do frame stacking to improve sensitivity (MDT), though the results depend greatly on how the camera generates the viewable format images. You can even get sub-pixel resolution by stacking slightly shifted images and doing appropriate signal processing. I am not aware of free or widely available software to do the latter two processes. You would have to search for the research papers on the processes and then write your own.
All of these things have their limitations. Any motion in the image, from leaves fluttering in the wind, clouds, water (waves, rivers, waterfalls, rain, sprinklers), and people and animals will spoil the effect. Also camera jitter can hurt, particularly if the motion occurs during the integration time of the sensor. I use similar techniques in the visible region and find that perhaps one in five times the image rivals what a better imager would generate. Another two times in five the resulting image is visually pleasing, but inaccurate in some way. The remainder are often literally unviewable. There is something wrong about them. You find them disturbing, but often can't pick out what is wrong.