A lot of photographs are collected in the cloud server through Smartphone based User Participating system such as Hiroshima Kankou Map Android application. The difficulties of analysis for the characteristic pattens related to the sightseeing spotsuare no measurement of similarity to the landmark. In this paper, the traditional image processing such as gray conversion,
Wavelet conversion, and the edge detection in OpenCV are implemented. We classified the encorded image of the collected photographs by SOM and report their results.

開催日：平成25年7月13日　会場：広島市立大学