作者: Ryszard Tadeusiewicz , Paweł Tylek , Florian Adamczyk , Paweł Kiełbasa , Mirosław Jabłoński
DOI: 10.3390/SU9122370
关键词:
摘要: Due to technological progress in forestry, seedlings with covered root systems—especially those grown container nurseries—have become increasingly important forest nursery production. One the trees that is most commonly this way common oak (Quercus robur L.). For an acorn be sown a container, it necessary remove its upper part during mechanical scarification, and evaluate sowing suitability. At present, mainly done manually by visual assessment. The low effectiveness of method preparation has encouraged search for unconventional solutions. them use automated device consists computer vision-based module. economic reasons related cost growing nurseries, beneficial minimize contribution unhealthy seeds. maximum accuracy, which understood as number correct seed diagnoses relative total seeds being assessed, was adopted criterion choosing separation threshold. According proposed, intensity red components images scarified acorns facilitated best results terms materials examined experiment. On average, 10% inaccuracy observed. A secondary outcome presented research evaluation ergonomic parameters user interface attached unit controlling when running autonomous operation mode.