This research project was proposed by ISAE-Supaero in partnership with Nuclétudes. The goal was to use CMOS Image sensors (CIS) to identify ionizing particles (alpha, X-rays, cosmic-rays).
This function calculates the standard deviations of the submatrices of a given matrix. Since the fixed pattern noise has a small standard deviation, this criterion can be used to distinguish between particle events and the former.
Retrives RAW data from the .JPG file.
Opens a .JPG image with a specified integration time (see images provided) and retrives the RAW data. The image processing algorithm is comprised of a Wiener filter followed by a convolution with an event kernel whose matrix is also provided. After these steps the frame obtained can be easily used for event detection using just a threshold or the subDeviations values.
Plots (main.m):
- Initial frame;
- Frame resulting from the wiener filter;
- The convoluted frame;
- Histogram of the average energy per event in eV;
- Representations of all the events identified.
NOTE: All the constants used in the calculations were determined and calibrated for a SONY IMX219 image sensor.
For more information please refer to the article provided: CIS_for_Particle_Detection_Bielawski_Li_Moreira.pdf
- Prof. Vincent Goiffon - ISAE Supaero (Image Sensor Research Team)
- Pierre Li
- Romain Bielawski
- Gabriel Moreira, 2018