engleski

Object Detection Based on Gray-scale Morphology

The paper describes an algorithm for object detection based on gray-scale morphology. Each channel of the color picture is separately processed with "black hat" or "top hat" morphological transform followed by thresholding. The threshold is decreased in each step of the algorithm until the expected percentage of black pixels in resulting binary image is reached. If the predefined minimal threshold is reached, the algorithm detects no objects in the image. The algorithm was tested on the detection of bar-code stickers on library books. The experimental results are satisfactory and encouraging for further research.

gray-scale morphology; object detection; top hat; black hat; barcode

nije evidentirano