Title :
Contribution of research & development to student´s motivation
Author :
Di Carlo, Paolo ; Barsics, Joseph ; Moitroux, Cecile
Author_Institution :
Dept. of Appl. Sci. & Comput., Haute Ecole de la Province de Liege, Liège, Belgium
Abstract :
A more than two year long and successful R&D project was carried out in partnership with another university and companies interested in the automation of aggregate size analysis in the mineral industry, with an overall objective to provide real-time feedback to the plant and to decrease energy consumption. The kernel of the project is a 3D vision system which exploits both CPU and GPU capabilities of personal computers to provide real time analysis and further processing. The system currently acquires one 1536 by 3072 pixels image per second. Images are processed and analysed in real-time to extract approximately 150000 particles every hour. A history of each computed parameter is available through a database where all of them are stored. The main challenge was optimization of the code which is now successful and the provided processes are nowadays in use and well running since several months. The main young software-developer researcher of the project was offered the opportunity to make a presentation of his work to a group of computer science (CS) students in the context of an image processing and industrial vision course (Master 2 level). They learned about application design, image processing tools and subtleties, code optimization, state-of-the-art GPUs and their parallel processors. The harmony and efficiency with which knowledge from distinct domains was applied in his work led students to adopt and successfully implement optimization techniques throughout their lab assignment. Compared to previous years, the student´s achievements were significantly and positively influenced by the R&D point of view added to the course. It´s also worth noting that computing students were particularly interested by the practical optimization concepts introduced during the presentation. Evidently, integrating theory from different courses into a single project presentation eliminates the singular difficulties and highlights the relevant features of each course. Ev- ry technical aspect of the R&D project and students achievements is described.
Keywords :
computer aided instruction; computer science education; computer vision; energy consumption; parallel processing; real-time systems; research and development; 3D vision system; CPU capabilities; CS students; GPU capabilities; R&D project; aggregate size analysis automation; application design; code optimization; computer science students; energy consumption; image pixels; image processing tools; industrial vision course; mineral industry; parallel processors; personal computers; real time analysis; real-time feedback; real-time image processing; research & development; student achievements; student motivation; subtleties; Computer architecture; Graphics processing units; Image processing; Optimization; Pipeline processing; Pipelines; cogolin; image processing; optimization; pipeline architecture; student´s motivation;
Conference_Titel :
EAEEIE Annual Conference (EAEEIE), 2013 Proceedings of the 24th
Conference_Location :
Chania
Print_ISBN :
978-1-4799-0042-8
DOI :
10.1109/EAEEIE.2013.6576499