Increasing signal-to-noise ratio of registered images by using light spatial noise portrait of camera's photosensor

Abstract

Increase of signal-to-noise ratio (SNR) of registered images is important task in such fields as image encryption, digital holography, pattern recognition, etc. Method of image SNR increasing by using light spatial noise portrait (LSNP) of camera's photosensor is presented. Use of proposed LSNP compensation method is especially effective after application of other methods for SNR increasing that suppress temporal noise. Procedure of LSNP measurement is described. LSNP of camera Canon EOS 400D was measured. Analytical expressions for estimation of achievable SNR increase were derived. Using characteristics of obtained LSNP, numerical experiments on estimation of SNR increase were performed. It is obtained that typically utilizing of averaging over frames method allows to increase SNR up to 2 times. Consequent application of LSNP compensation method leads to 10 times SNR increase. These numerical experiments results confirm derived analytical expressions. In case of using more accurate LSNP compared to the obtained, SNR can be increased up to 50 times. Using cameras Canon EOS 400D and MegaPlus II ES11000, test experiments were performed. Experimental results are in good agreement with numerical ones. Experimentally obtained SNR increased 5 times compared to the original.