We used two methods of image processing to maximise the information from our observations...
The "Cheap 'n Dirty" method
Using Paintshop Pro V5 and applying it's standard algorithms
The "Hi-Tech" method - Maximum Entropy Deconvolution
This technique uses information from multiple frames on the captured video to reduce background noise on the image and then sharpen the image.
First we selected the four best frames from the raw video, these are all less than a second apart
Next these images were aligned and superimposed into a single image using a PC based Maths package which manipulates the images as arrays of 8-bit colour information. This produces the base image in the following table. Note the reduction in background noise compared with a raw frame
Finally a de-convolution algorithm was applied to the base image using a using a 5 pixel square sharpen point spread function. Then we used the techniques described above to soften and contrast stretch the image. Using a soften function after de-convolution seems counter-intuitive but it does result in a more pleasing image!
Astrostack - Freeware
Since writing the above page where the raw images were processed using the software and algorithm which Phil developed we have now processed the same frames through the Astrostack software which produced a very similar result - highly recommended for anybody starting out to use these techniques.
Comparing the final images...
Note the following in the processed image:-
Images and content ©2001 Philip Masding & Mike