The　paper　first　expounds　the　principle　of　synergetic　pattern　recognition.　Then　it　discusses　the　implementation　techniques　of　the,pathology　image　classification　system　based　on　the　principle　in　detail.　What＇s　more,　to　resolve　the　problem　that　the　synergetic　neural　network　is　weak　in　recognizing　new　test　samples,　it　presents　the　information　superposition　self-learning　algorithm　based　on　genetic　algorithm.　The　experimental　results　prove　the　effectivity　of　the　algortihm.