We tested the applicability of the sonogram detector and a rule-based evaluation for seismic source classification and identification in Israel. Source identification is based on the transformation of full seismograms to images and the correlation between a new image and known images from different seismic sources. These methods were initially developed to monitor the induced seismicity of northwest Germany by the BUG small array. Two new elements were introduced in this study: (a) the adaptation of the array technique to the identification of seismic sources using a local network and (b) the recognition of source regions with natural seismicity. The tested data include 41 local earthquakes and explosions in the time period September 1997 to December 1997. Automated processing yielded 95% success in correctly identifying the individual sources with no misclassification. One event was not classified. Path and distance effects did not influence the classification capability. The method could be useful for seismic verification under the Comprehensive Test Ban Treaty.