Food forgery is one of the most articulated socio-economic concerns, which contributed to increase people awareness on what they eat. Identification of species represents a key aspect to expose commercial frauds implemented by substitution of valuable species with others of lower value. Fish species identification is mainly performed by morphological identification of gross anatomical features of the whole fish. However, the increasing presence on markets of new little-known species makes morphological identification of species difficult. In this paper we present FishAPP, a cloud-based infrastructure for fish species recognition. FishAPP is composed of a mobile application developed for the Android and the iOS mobile operating system enabling the user to shot pictures of a whole fish and submit them for remote analysis and a remote cloud-based processing system that implements a complex image processing pipeline and a neural network machine learning system able to analyze the obtained images and to perform classification into predefined fish classes. Preliminary results obtained from the available dataset provided encouraged results.

FishAPP: A mobile App to detect fish falsification through image processing and machine learning techniques / Rossi, Francesco; Benso, Alfredo; DI CARLO, Stefano; Politano, GIANFRANCO MICHELE MARIA; Savino, Alessandro; Acutis, Pier Luigi. - ELETTRONICO. - (2016), pp. 1-6. (Intervento presentato al convegno 20th IEEE International Conference on Automation, Quality and Testing, Robotics, AQTR 2016 tenutosi a Cluj-Napoca, RO nel 19-21 May 2016) [10.1109/AQTR.2016.7501348].

FishAPP: A mobile App to detect fish falsification through image processing and machine learning techniques

ROSSI, FRANCESCO;BENSO, ALFREDO;DI CARLO, STEFANO;POLITANO, GIANFRANCO MICHELE MARIA;SAVINO, ALESSANDRO;
2016

Abstract

Food forgery is one of the most articulated socio-economic concerns, which contributed to increase people awareness on what they eat. Identification of species represents a key aspect to expose commercial frauds implemented by substitution of valuable species with others of lower value. Fish species identification is mainly performed by morphological identification of gross anatomical features of the whole fish. However, the increasing presence on markets of new little-known species makes morphological identification of species difficult. In this paper we present FishAPP, a cloud-based infrastructure for fish species recognition. FishAPP is composed of a mobile application developed for the Android and the iOS mobile operating system enabling the user to shot pictures of a whole fish and submit them for remote analysis and a remote cloud-based processing system that implements a complex image processing pipeline and a neural network machine learning system able to analyze the obtained images and to perform classification into predefined fish classes. Preliminary results obtained from the available dataset provided encouraged results.
2016
9781467386906
File in questo prodotto:
File Dimensione Formato  
AQTR16_Fishapp_AuthorVersion.pdf

accesso aperto

Descrizione: Author version
Tipologia: 2. Post-print / Author's Accepted Manuscript
Licenza: PUBBLICO - Tutti i diritti riservati
Dimensione 1.8 MB
Formato Adobe PDF
1.8 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2649750
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo