Computer vision based street width measurement for urban aesthetics identification

Abstract

This paper presents a computer-vision based Street-Width measurement system for Urban Aesthetics identification. In this system image is captured using a digital camera. In the pre-processing section the image is scaled in high to low resolution. The methodology used to identify object was contour tracing and canny edge detection algorithm. Then the object is measured by generalizing the pixel mapping. Finally the finding output is matched with the Standard Street Measurement. The measurement findings in this proposed methodology was then analyzed and based on some criteria the street aesthetics of Dhaka city was provided.