Title: Image Processing and hyperspectral imaging for quality control in the potato industry
Student: Angel Dacal Nieto
PhD advisers: Dr. Arno Formella and Dr. Pilar Carrión Pardo
Presentation: July 7, 2011.
My pre-doc stage began on 2007, with the PhD program "Advanced Technologies in Intelligent Software Development" of the Computer Science Department of Universidade de Vigo. After complete the 2007/2009 course, I acquired the Diploma de Estudios Avanzados (DEA), on July 2009, in the speciality of Image Processing.
Then, I wrote my PhD dissertation called "Procesamiento de imagen y visión hiperespectral para el control de calidad en la industria de la patata (Image Processing and hyperspectral imaging for quality control in the potato industry)", collaborating with the group Laboratorio de Informática Aplicada (LIA), and my advisers Dr. Arno Formella and Dr. Pilar Carrión Pardo.
In July 7, 2011, I presented my dissertation. The Thesis committee was composed by Dr. Filiberto Pla (Universitat Jaume I), Dr. Pablo García (Universidad de Extremadura), Dr. Pedro Villar (Universidad de Granada), Dr. Eva Cernadas (Universidade de Santiago de Compostela) and Dr. María José Lado (Universidade de Vigo). The committee assigned me the highest qualification: Sobresaliente Cum Laude.
The overall objective of this dissertation was the design and analysis of image processing and hyperspectral imaging techniques to automate certain tasks relating to quality control in the potato industry which are developed manually until now. This study covered the classical stages of computer vision: design and implementation of an image acquisition system, preprocessing and segmentation through various image processing techniques, feature extraction and classification.
Some of the tasks that have been developed are:
- Classification of tubers depending on their external defects (greening, rotten, etc.).
- Detecting hollow heart in potatoes using hyperspectral imaging.
- Estimating area affected by common scab in potatoes using hyperspectral imaging.
- Camera JAI BB-500GE (sensor CCD color 2/3", 2456x2058, 15fps, GigE Vision) and lens Schneider Cinegon Xenoplan 1.4/17mm 2/3" series Compact 400-1000.
- Hypespectral imaging NIR system
- Illumination by fiber source Schott DCR III Plus, and randomized bundle.
- Camera JAI BM-500GE (sensor CCD monochrome 2/3", 2456x2058, 15fps, GigE Vision)
- Camera JAI CB-200GE (sensor CCD color 1/1.8", 1620x1236, 25fps, GigE Vision)
- Camera C-Cam BCi4-USB (sensor CMOS color , 1280x1024, 14fps, USB)
- Camera Thorlabs DC210C (sensor CCD, 640x480, FireWire)
Used libraries and technologies:
- JAI SDK
- XControl API
- MCode Intelligent Motion Systems language
Derived publications (until now):
To be published in IEEE Xplore. 2011.
Presented at ICME 2011 (IEEE International Conference on Multimedia and Expo). Barcelona (Spain), from11-07-2011 to15-07-2011.
Article in Industrial Electronics, 2009. IECON'09. 35th Annual Conference of IEEE, pp. 1955-1960 2009. ISBN: 978-1-4244-4648-3.
Presented in IECON 2009 (35th Internation Annual Conference of the IEEE Industrial Electronics Society). Porto (Portugal), from 03-11-2009 to 05-11-2009. Presentation.