遺伝的アルゴリズム及びニューラルネットワークを用いた自然画像からの農産物の認識

概要

論文の詳細を見る
The harvesting operation is a hard work for farmers. Workers hope the reduction of the harvesting operation in the agriculture. Then, the purpose of this study is the construction of useful visual recognition system that requires in automation and robotization of harvesting operation. In this research, eggplants and tomatoes are taken up as an example of agricultural objects. First, these are searched using genetic algorithm from scene image. second, the recognition method of agricultural objects using the Neural-Networks is proposed. In this Neural-Networks, the change of luminance information has been learned as the feature of agricultural objects, which shows a luminosity distribution for the object composed of curved surface.Finally, through experimental studies, it is shown that new technology proposed here is sufficiently valid on recognition for agricultural object.