Fast Event-based Corner Detection

Abstract

Event cameras offer many advantages over standard frame-based cameras, such as low latency, high temporal resolution, and a high dynamic range. They respond to pixel-level brightness changes and, therefore, provide a sparse output. However, in textured scenes with rapid motion, millions of events are generated per second. Therefore, state-of-the-art event-based algorithms either require massive parallel computation (e.g., a GPU) or depart from the event-based processing paradigm. Inspired by frame-based pre-processing techniques that reduce an image to a set of features, which are typically the input to higherlevel algorithms, we propose a method to reduce an event stream to a corner event stream. Our goal is twofold: extract relevant tracking information (corners do not suffer from the aperture problem) and decrease the event rate for later processing stages. Our event-based corner detector is very efficient due to its design principle, which consists of working on the Surface of Active Events (a map with the timestamp of the latest event at each pixel) using only comparison operations. Our method asynchronously processes event by event with very low latency. Our implementation is capable of processing millions of events per second on a single core (less than a micro-second per event) and reduces the event rate by a factor of 10 to 20. AbstractEvent cameras offer many advantages over standard frame-based cameras, such as low latency, high temporal resolution, and a high dynamic range. They respond to pixellevel brightness changes and, therefore, provide a sparse output. However, in textured scenes with rapid motion, millions of events are generated per second. Therefore, stateof-the-art event-based algorithms either require massive parallel computation (e.g., a GPU) or depart from the event-based processing paradigm. Inspired by frame-based pre-processing techniques that reduce an image to a set of features, which are typically the input to higher-level algorithms, we propose a method to reduce an event stream to a corner event stream. Our goal is twofold: extract relevant tracking information (corners do not suffer from the aperture problem) and decrease the event rate for later processing stages. Our event-based corner detector is very efficient due to its design principle, which consists of working on the Surface of Active Events (a map with the timestamp of the latest event at each pixel) using only comparison operations. Our method asynchronously processes event by event with very low latency. Our implementation is capable of processing millions of events per second on a single core (less than a micro-second per event) and reduces the event rate by a factor of 10 to 20.

FA-Harris: A Fast and Asynchronous Corner Detector for Event Cameras

“…Inspired by frame-based corner detection techniques, Mueggler et al [20] proposed a FAST-like event-based corner detection method which is called eFAST. They detected the event-corners on SAE and reduced event-streams to event-corner streams.…”

Speed Invariant Time Surface for Learning to Detect Corner Points With Event-Based Cameras

“…4.2) is a common representation used in event-based vision [7,39,65,55]. It has been used in [39] for corner detection, where the FAST algorithm [50] is adapted to the pattern of Time Surface of a moving corner. In [4] this formulation has been extended to corners with obtuse angles.…”

“…One of the main fundamental tasks is feature point (a) Events (b) evFAST [39] (c) evHarris [61] (d) Ours Figure 1: (a) Stream of events generated by an event-based camera moving in front of a checkerboard pattern. Black dots represents events with a negative polarity, white dots events with a positive polarity.…”

“…Inspired by the vast literature on frame-based feature point detection, some works adapted frame-based corner detector to event-based data. Typically, a local spatial descriptor is built around an event, for example by cumulating events in a given time window [61], or by considering the times of arrival of the events [39]. Then, a classical test, such as [26] can be applied to this 2D spatial neighborhood.…”

Neuromorphic vision: From sensors to event‐based algorithms

“…The computational complexity of the previous methods was overcome in Mueggler, Bartolozzi, and Scaramuzza (), where a corner detection method inspired by conventional FAST corner detector (Rosten & Drummond, ) has been proposed. The method operated on Surface of Active Events (SAE) (Benosman, Clercq, Lagorce, Ieng, & Bartolozzi, ).…”