Shapes Characterization on Address Event Representation Using Histograms of Oriented Events and an Extended LBP Approach

TL;DR

This paper introduces a novel shape characterization method using Address Event Representation with histograms of oriented events and an extended LBP approach, achieving high accuracy in symbol and digit recognition tasks.

Contribution

It presents a new descriptor for shape analysis in event-based data and demonstrates its effectiveness with simple classifiers on challenging datasets.

Findings

Achieves 98.5% accuracy on Poker-DVS dataset.

Attains 94.6% and 96.3% accuracy on MNIST-DVS digits.

Outperforms some state-of-the-art methods with limited events.

Abstract

Address Event Representation is a thriving technology that could change digital image processing paradigm. This paper proposes a methodology to characterize the shape of objects using the streaming of asynchronous events. A new descriptor that enhances spikes connectivity is associated with two oriented histogram based representations. This paper uses these features to develop both a non-supervised and a supervised multi-classification framework to recognize poker symbols from the Poker-DVS public dataset. The aforementioned framework, which uses a very limited number of events and a simple class modeling, yields results that challenge more sophisticated methodologies proposed by the state of the art. A feature family based on context shapes is applied to the more challenging 2015 Poker-DVS dataset with a supervised classifier obtaining an accuracy of 98.5 %. The system is also applied to the MNIST-DVS dataset yielding an accuracy of 94.6 % and 96.3 % on digit recognition, for scales 4 and 8 respectively.