Open AccessEstimation of edge displacement against brightness and camera‐to‐object distance
Author(s) -
Seo Suyoung
Publication year - 2017
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/iet-ipr.2016.0796
Subject(s) - brightness , subpixel rendering , ground truth , enhanced data rates for gsm evolution , artificial intelligence , displacement (psychology) , computer vision , weighting , pixel , contrast (vision) , edge detection , mathematics , computer science , image (mathematics) , optics , physics , image processing , acoustics , psychology , psychotherapist
This study proposes a systematic method to estimate edge displacement (ED) from ground truth edges. To obtain ground truth edges and test edges, a series of target sheets containing the reference regions and brightness contrast regions were designed. Further, the brightness contrast regions were designed to contain foreground and background brightness between which the image edges occur. By varying the foreground brightness (FB), the influence of the FB on the edge locations was tested. Moreover, the influence of camera‐to‐object distance on the edge locations was tested. A simple least‐squares method using slope profiles and a slope‐based weighting scheme is proposed for calculating the edges occurring in contrast regions with subpixel accuracy. The experiment results revealed that a bright foreground moves the edge location towards the dark side of the edges in the range of 0–0.6 pixels with strong relationships between the FB and magnitude of ED.