(Akira Okumura, Yasunobu Uchiyama, Tsunefumi Mizuno, Johann Cohen-Tanugi)
We have developed an image restoration technique based on the Richardson-Lucy algorithm optimized for Fermi-LAT image analysis. This technique is unique since it takes advantage of the PSF class of the Pass8 data set for each event, which is critical for Fermi-LAT image analysis since the PSF depends on the energy and angle of incident gamma rays and varies by more than one order of magnitude. Fermi-LAT image analysis also faces Poisson noise due to low photon statistics. Our technique incorporates wavelet filtering to minimize such noise effects. In addition, this technique can utilize the diffuse background model as a part of the prior image so that we can produce an extended image of residual gamma rays. We present studies of galactic extended sources and point sources in the galactic center region using this image restoration technique.