Detection of grapes in natural environment using HOG features in low resolution images
Konferenční objektOtevřený přístuppeer-reviewedpostprintSoubory
Datum publikování
2017
Autoři
Vedoucí práce
Oponent
Název časopisu
Název svazku
Vydavatel
Institute of Physics Publishing Ltd
Abstrakt
Detection of grapes in real-life images has importance in various viticulture applications. A grape detector based on an SVM classifier, in combination with a HOG descriptor, has proven to be very efficient in detection of white varieties in high-resolution images. Nevertheless, the high time complexity of such utilization was not suitable for its real-time applications, even when a detector of a simplified structure was used. Thus, we examined possibilities of the simplified version application on images of lower resolutions. For this purpose, we designed a method aimed at search for a detector’s setting which gives the best time complexity vs. performance ratio. In order to provide precise evaluation results, we formed new extended datasets. We discovered that even applied on low-resolution images, the simplified detector, with an appropriate setting of all tuneable parameters, was competitive with other state of the art solutions. We concluded that the detector is qualified for real-time detection of grapes in real-life images.
Rozsah stran
p. 1-8
ISSN
1742-6588
Trvalý odkaz na tento záznam
Projekt
SGS_2017_027/Moderní metody simulace, řízení a optimalizace
Zdrojový dokument
Journal of Physics: Conference Series
Vydavatelská verze
Přístup k e-verzi
open access
Název akce
2nd International Conference on Measurement Instrumentation and Electronics, ICMIE 2017 (09.06.2017 - 11.06.2017, Praha)
ISBN
Studijní obor
Studijní program
Signatura tištěné verze
Umístění tištěné verze
Přístup k tištěné verzi
Klíčová slova
grape detection, precision viticulture, real scene images, image processing, histogram of oriented gradients, support vector machine