We consider a new application paradigm of vehicular sensor networks (VSN). Currently, vehicles are equipped with forward facing cameras to assist forensic investigations of events by proactive image-capturing from streets and roads. Due to content redundancy and storage imbalance in this in-network