A DSEL for High Throughput and Low Latency Software-Defined Radio on Multicore CPUs

Cassagne, Adrien; Tajan, Romain; Aumage, Olivier; Leroux, Camille; Barthou, Denis; Jégo, Christophe

doi:10.48550/arxiv.2206.06147

Cited by 1 publication

(1 citation statement)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, the minimum time required for a manual detection is close to the full time of the video sequence while the proposed application is real-time and compatible with the cubesat power consumption constraint. Moreover, FMDT is able to leverage multi-core processor architectures through a task graph description and the use of the AFF3CT multi-threaded runtime library (Cassagne et al 2019(Cassagne et al , 2021(Cassagne et al , 2022. AFF3CT was designed for digital communication systems but is well adapted to real-time image processing.…”

Section: Computer-vision Detectionmentioning

confidence: 99%

A 2022τ-Herculid meteor cluster from an airborne experiment: automated detection, characterization, and consequences for meteoroids

Vaubaillon

Loir

C.³

et al. 2023

A&A

View full text Add to dashboard Cite

Context. The existence of meteor clusters has long since been a subject of speculation and so far only seven events have been reported, among which two involve less than five meteors, and three were seen during the Leonid storms. Aims. The 1995 outburst of Comet 73P/Schwassmann-Wachmann was predicted to result in a meteor shower in May 2022. We detected the shower, proved this to be the result of this outburst, and detected another meteor cluster during the same observation mission. Methods. The τ-Herculids meteor shower outburst on 31 May 2022 was continuously monitored for 4 h during an airborne campaign. The video data were analyzed using a recently developed computer-vision processing chain for meteor real-time detection. Results. We report and characterize the detection of a meteor cluster involving 38 fragments, detected at 06:48 UT for a total duration of 11.3 s. The derived cumulative size frequency distribution index is relatively shallow: s = 3.1. Our open-source computer-vision processing chain (named FMDT) detects 100% of the meteors that a human eye is able to detect in the video. Classical automated motion detection assuming a static camera was not suitable for the stabilized camera setup because of residual motion. Conclusions. From all reported meteor clusters, we crudely estimate their occurrence to be less than one per million observed meteors. Low heliocentric distance enhances the probability of such meteoroid self-disruption in the interplanetary space.

show abstract

Section: Computer-vision Detectionmentioning

confidence: 99%