Abstract
We initiate a systematic study of tolerant testers of image properties or, equivalently, algorithms that approximate the distance from a given image to the desired property (that is, the smallest fraction of pixels that need to change in the image to ensure that the image satisfies the desired property). Image processing is a particularly compelling area of applications for sublineartime algorithms and, specifically, property testing. However, for testing algorithms to reach their full potential in image processing, they have to be tolerant, which allows them to be resilient to noise. Prior to this work, only one tolerant testing algorithm for an image property (image partitioning) has been published.
We design efficient approximation algorithms for the following fundamental questions: What fraction of pixels have to be changed in an image so that it becomes a halfplane? a representation of a convex object? a representation of a connected object? More precisely, our algorithms approximate the distance to three basic properties (being a halfplane, convexity, and connectedness) within a small additive error epsilon, after reading a number of pixels polynomial in 1/epsilon and independent of the size of the image. The running time of the testers for halfplane and convexity is also polynomial in 1/epsilon. Tolerant testers for these three properties were not investigated previously. For convexity and connectedness, even the existence of distance approximation algorithms with query complexity independent of the input size is not implied by previous work. (It does not follow from the VCdimension bounds, since VC dimension of convexity and connectedness, even in two dimensions, depends on the input size. It also does not follow from the existence of nontolerant testers.)
Our algorithms require very simple access to the input: uniform random samples for the halfplane property and convexity, and samples from uniformly random blocks for connectedness. However, the analysis of the algorithms, especially for convexity, requires many geometric and combinatorial insights. For example, in the analysis of the algorithm for convexity, we define a set of reference polygons P_{epsilon} such that (1) every convex image has a nearby polygon in P_{epsilon} and (2) one can use dynamic programming to quickly compute the smallest empirical distance to a polygon in P_{epsilon}. This construction might be of independent interest.
BibTeX  Entry
@InProceedings{berman_et_al:LIPIcs:2016:6195,
author = {Piotr Berman and Meiram Murzabulatov and Sofya Raskhodnikova},
title = {{Tolerant Testers of Image Properties}},
booktitle = {43rd International Colloquium on Automata, Languages, and Programming (ICALP 2016)},
pages = {90:190:14},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
ISBN = {9783959770132},
ISSN = {18688969},
year = {2016},
volume = {55},
editor = {Ioannis Chatzigiannakis and Michael Mitzenmacher and Yuval Rabani and Davide Sangiorgi},
publisher = {Schloss DagstuhlLeibnizZentrum fuer Informatik},
address = {Dagstuhl, Germany},
URL = {http://drops.dagstuhl.de/opus/volltexte/2016/6195},
URN = {urn:nbn:de:0030drops61959},
doi = {10.4230/LIPIcs.ICALP.2016.90},
annote = {Keywords: Computational geometry, convexity, halfplane, connectedness, propertytesting, tolerant property testing}
}
Keywords: 

Computational geometry, convexity, halfplane, connectedness, propertytesting, tolerant property testing 
Collection: 

43rd International Colloquium on Automata, Languages, and Programming (ICALP 2016) 
Issue Date: 

2016 
Date of publication: 

23.08.2016 