In this paper we present the distributed event localization and tracking algorithm DELTA that solely depends on light measurements. Based on this information and the positions of the sensors, DELTA is able to track a moving person equipped with a flashlight by dynamically building groups and electing well located nodes as group leaders. Moreover, DELTA supports object localization. The gathered data is sent to a monitoring entity in a fixed network which can apply pattern recognition techniques to determine the legitimacy of the moving person. DELTA enables object tracking with minimal constraints on both sensor hardware and the moving object. We show the feasibility of the algorithm running on the limited hardware of an existing sensor platform. Keywords Sensor network, monitoring, tracking, hardware