Introducing a Decision Support System for Use-Case Specific Object Detection Methods in Production Systems

Download statistics - Document (COUNTER):

Müller, K.; Saß, S.-A.; Greb, C.: Introducing a Decision Support System for Use-Case Specific Object Detection Methods in Production Systems. In: Herberger, D.; Hübner, M. (Eds.): Proceedings of the Conference on Production Systems and Logistics : CPSL 2021. Hannover : publish-Ing., 2021, S. 605-615. DOI: https://doi.org/10.15488/11248

Selected time period:

year: 
month: 

Sum total of downloads: 63




Thumbnail
Abstract: 
Object detection has the potential to facilitate the automation of optical quality inspection, achieving asignificant reduction of human error. However, the gap between needed expertise to understand and integratecomplex object detection systems into production environments and the availability of computer scientists,is hindering its use in the manufacturing industry. A support system for decision-makers unrelated to thesubject is therefore required to promote industry utilisation of object detection and effectively manage thisotherwise unused opportunity of profitable knowledge. Within the Cluster of Excellency “Internet ofProduction (IoP)”, such a support system has been developed. Lowering the implementation hurdle of objectdetection systems is achieved by translating complex information about existing methods into tangiblefactors, such as quality and cost. In this work we aim to structure relevant object detection techniques andemploy a decision tree to provide a user-support based on a use-case-oriented framework. The use-case’sbasic conditions and requirements serve as input for the framework. The decision tree gives the suitableobject detection method as output, accordingly. For traditional object detection techniques, thecharacteristics are translated into basic requirements, which are then used as input, e.g., for the templatematching method the comparison of a source image with a reference image is translated to the ability toguarantee images for all possible quality deviations. The deep learning (DL) methods are consolidated in theproject management triangle consisting of quality, cost, and time. Firstly, an introduction into objectdetection is given. Secondly, traditional methods are clustered and deep learning methods classified. Adescription of the decision tree is then presented, before testing results conclude this paper. The developedsupport system enables decision-makers to evaluate object detection methods for individual use-cases andconsequently achieve increased production planning efficiency. The system’s universal design allows forapplication across manufacturing industries and use-cases.
License of this version: CC BY 3.0 DE
Document Type: bookPart
Publishing status: publishedVersion
Issue Date: 2021
Appears in Collections:Proceedings CPSL 2021
Proceedings CPSL 2021

distribution of downloads over the selected time period:

downloads by country:

pos. country downloads
total perc.
1 image of flag of Germany Germany 32 50.79%
2 image of flag of United States United States 6 9.52%
3 image of flag of Russian Federation Russian Federation 5 7.94%
4 image of flag of Japan Japan 4 6.35%
5 image of flag of Indonesia Indonesia 4 6.35%
6 image of flag of Czech Republic Czech Republic 3 4.76%
7 image of flag of Portugal Portugal 2 3.17%
8 image of flag of Hong Kong Hong Kong 2 3.17%
9 image of flag of China China 2 3.17%
10 image of flag of Iran, Islamic Republic of Iran, Islamic Republic of 1 1.59%
    other countries 2 3.17%

Further download figures and rankings:


Hinweis

Zur Erhebung der Downloadstatistiken kommen entsprechend dem „COUNTER Code of Practice for e-Resources“ international anerkannte Regeln und Normen zur Anwendung. COUNTER ist eine internationale Non-Profit-Organisation, in der Bibliotheksverbände, Datenbankanbieter und Verlage gemeinsam an Standards zur Erhebung, Speicherung und Verarbeitung von Nutzungsdaten elektronischer Ressourcen arbeiten, welche so Objektivität und Vergleichbarkeit gewährleisten sollen. Es werden hierbei ausschließlich Zugriffe auf die entsprechenden Volltexte ausgewertet, keine Aufrufe der Website an sich.

Search the repository


Browse