SEGMENTATION OF DIGITAL IMAGES FROM A ROAD TRAFFIC VIDEORECORDER
Zbigniew Czapla · 2012
Summary. The proposed segmentation method uses the two-layer data model of digital images. The two-layer data model consists of a different value layer and a base value layer. An original image pixel values are obtained by appropriate combining pixel values of both layers. A base pixel value together with corresponding difference pixel values constitute a data cluster. The principle of segmentation is division of an image greyscale into disjoint ranges and assigning of individually base values to the proper ranges. The segmentation process utilizes the base values mainly and therefore proposed technique is computationally attractive.