Object Isolation Using a Pulse-Coupled Neural Network

ABSTRACT Object isolation in an image is an iterative process in which the target (object beginsought) is gradually enhanced and the non-targets (all other objects within the scene)gradually fade. The intent ofthis design is to be more robust to noise, object alterations, and clutter than traditional ifiters. This iterative process can be performed by a non-linear shunting Pulse-Coupled Neural Network (PCNN). The input scene is modified by the filtered pulse response ofthe PCNN. Within several iterations (less than 50) the target isenhanced and the non-targets are degraded to such a degree that the target is completely dominant in the input. The segmented pulse response of the PCNN allows for a uniqueapproach to this system.1. THEORYThe Pulse-Coupled Neural Network (PCNN) [Johnson,94] is a cortical model basedupon the biological findings ofEckhorn [90]. The PCNN has many inherent features whichincludes two very important features for image processing: segmentation and edgeextraction. The PCNN creates a series ofbinary pulse images. Each ofthese contain pulseswhich display a subset ofobjects that are within the input scene. This segmentation ability