Some FPGA FIR resources :
https://www.fpga4student.com/2016/11/image-processing-on-fpga-verilog.html
https://www.dspguide.com/ch6/3.htm
https://www.hackster.io/whitney-knitter/dsp-for-fpga-simple-fir-filter-in-verilog-91208d
https://github.com/samiyaalizaidi/FIR-Filter
On DSP more generally :
https://www.analog.com/en/resources/technical-books/scientist_engineers_guide.html
https://visionbook.mit.edu/
On the rpi camera :
https://picamera.readthedocs.io/en/release-1.13/api_camera.html
****
Here’s the FIR filter output using samiya alizaidi’s code lightly modified (to take in only 2 bits and output 3 bits) :
Next step is to be able to turn it on and off with a button to see the effect better, and to have a potentiometer (or even a new board with a bunch of bots) to vary the taps live somehow.
****
I want to convolute in place an image recorded in BRAM. This will be a combination of down sampling and convoluting in parallel.
*****
MIT class using the OV7670 camera and an iCE40 :
https://github.com/westonb/OV7670-Verilog
Another MIT class which has a microchip to initialize the camera !
https://fpga.mit.edu/6205/F22/documentation/camera_board
The camera’s datasheet :
https://web.mit.edu/6.111/www/f2016/tools/OV7670_2006.pdf
Sobel filter implementation :
https://www.cs.columbia.edu/~sedwards/classes/2022/4840-spring/reports/CameraControl-report.pdf
opening up the rpi camera and trying different lenses :