Visible/Near Infrared Fusion
Configured the H7 to be able to load in an image downsampled by a factor of 9
From 564x747 downsampled to 63x83, seems like the best we can do because of the fundamental limits of the H7, this is also 3 Channel as well
I increased the heap and stack size that the H7 will use for this project. I also modified the image loading code to allocated up to 32Kb of memory for the incoming image
I have also made sure the H7 loads in a 3 channel image as a flattened array in that each channel is just concatenated onto the previous
We are now ready to implement the Fusion algorithm on the H7, it has been complete in MATLAB
I spoke with Jiawei about how this would affect our results being that our project is based on recognizing a difference in detail after the fusion of the image, he said that it is okay if we dont produce the best results as long as the algorithm is implemented well on the H7
OK, so I believe that you are the first team to be able to download a 3-channel RGB image of any size onto the H7. IIRC, there is at least one other team that is trying to do the same thing.
I agree with you and Jiawei that showing that the image processing actually works on the H7 is sufficient. We definitely want to see the Matlab results as well.