Detecting motorcycle helmet use with deep learning
| dc.contributor.author | Siebert, Felix Wilhelm | |
| dc.contributor.author | Lin, Hanhe | |
| dc.date.accessioned | 2019-11-12T16:50:17Z | |
| dc.date.available | 2019-11-12T16:50:17Z | |
| dc.date.issued | 2019-11-06 | |
| dc.description.abstract | The continuous motorization of traffic has led to a sustained increase in the global number of road related fatalities and injuries. To counter this, governments are focusing on enforcing safe and law-abiding behavior in traffic. However, especially in developing countries where the motorcycle is the main form of transportation, there is a lack of comprehensive data on the safety-critical behavioral metric of motorcycle helmet use. This lack of data prohibits targeted enforcement and education campaigns which are crucial for injury prevention. Hence, we have developed an algorithm for the automated registration of motorcycle helmet usage from video data, using a deep learning approach. Based on 91,000 annotated frames of video data, collected at multiple observation sites in 7 cities across the country of Myanmar, we trained our algorithm to detect active motorcycles, the number and position of riders on the motorcycle, as well as their helmet use. An analysis of the algorithm's accuracy on an annotated test data set, and a comparison to available human-registered helmet use data reveals a high accuracy of our approach. Our algorithm registers motorcycle helmet use rates with an accuracy of −4.4% and +2.1% in comparison to a human observer, with minimal training for individual observation sites. Without observation site specific training, the accuracy of helmet use detection decreases slightly, depending on a number of factors. Our approach can be implemented in existing roadside traffic surveillance infrastructure and can facilitate targeted data-driven injury prevention campaigns with real-time speed. Implications of the proposed method, as well as measures that can further improve detection accuracy are discussed. | en |
| dc.identifier.eissn | 1879-2057 | |
| dc.identifier.issn | 0001-4575 | |
| dc.identifier.uri | https://depositonce.tu-berlin.de/handle/11303/10288 | |
| dc.identifier.uri | http://dx.doi.org/10.14279/depositonce-9250 | |
| dc.language.iso | en | en |
| dc.relation.issupplementedby | http://dx.doi.org/10.14279/depositonce-9259 | en |
| dc.rights.uri | https://creativecommons.org/licenses/by-nc-nd/4.0/ | en |
| dc.subject.ddc | 006 Spezielle Computerverfahren | de |
| dc.subject.ddc | 150 Psychologie | de |
| dc.subject.ddc | 380 Handel, Kommunikation, Verkehr | de |
| dc.subject.other | deep learning | en |
| dc.subject.other | helmet use detection | en |
| dc.subject.other | motorcycle | en |
| dc.subject.other | road safety | en |
| dc.subject.other | injury prevention | en |
| dc.title | Detecting motorcycle helmet use with deep learning | en |
| dc.type | Article | en |
| dc.type.version | acceptedVersion | en |
| dcterms.bibliographicCitation.articlenumber | 105319 | en |
| dcterms.bibliographicCitation.doi | 10.1016/j.aap.2019.105319 | en |
| dcterms.bibliographicCitation.journaltitle | Accident Analysis & Prevention | en |
| dcterms.bibliographicCitation.originalpublishername | Elsevier | en |
| dcterms.bibliographicCitation.originalpublisherplace | Amsterdam [u.a.] | en |
| dcterms.bibliographicCitation.volume | 134 | en |
| tub.accessrights.dnb | free | en |
| tub.affiliation | Fak. 5 Verkehrs- und Maschinensysteme::Inst. Psychologie und Arbeitswissenschaft::FG Arbeits-, Ingenieur- und Organisationspsychologie | de |
| tub.affiliation.faculty | Fak. 5 Verkehrs- und Maschinensysteme | de |
| tub.affiliation.group | FG Arbeits-, Ingenieur- und Organisationspsychologie | de |
| tub.affiliation.institute | Inst. Psychologie und Arbeitswissenschaft | de |
| tub.publisher.universityorinstitution | Technische Universität Berlin | en |