書誌事項
- タイトル別名
-
- Robotic Harvesting System for Eggplants Trained in V-Shape (Part 1)-System Composition and Fruit Detection-
- Vジガタ セイシ サレタ ナス オ タイショウ ト シタ ロボット シュウカク システム ダイ1ポウ システム コウセイ ト カジツ ケンシュツ
- System Composition and Fruit Detection
- システム構成と果実検出
この論文をさがす
抄録
A robotic harvesting system was developed to realize selective harvesting for eggplants (Solanum melongena L.) trained in V-shape in a greenhouse. The purpose of the first part of this study is to explain the system components and the global sensing for detecting fruit and to examine the image scanning. The system was composed of control unit, sensing unit, manipulator unit, end-effector unit, and traveling vehicle. The system could run at an inter-ridge space and scan the image of eggplants set at both sides with a combination of the vehicle traveling and manipulator control. The global sensing could detect the existence of fruit from approximately 500 mm away using an image processing technique. The basic experiment of image scanning was conducted to evaluate the global sensing. The success detection ratio was 69.3%. However, unsuccessful detections were observed, such as fruit on an adjacent plant, under-sized fruit, and shade. The scanning time, which includes traveling, image capture, and manipulator motion, within 4 m in length and 1.7 m in height was 317 s. The manipulator motion occupied most of the total scanning time. The fruit direction could be calculated from the centroids of the detected fruit.
収録刊行物
-
- 植物工場学会誌
-
植物工場学会誌 15 (4), 205-210, 2003
日本生物環境工学会
- Tweet
詳細情報 詳細情報について
-
- CRID
- 1390001204380791808
-
- NII論文ID
- 10012024885
-
- NII書誌ID
- AN10451761
-
- ISSN
- 18803555
- 09186638
-
- NDL書誌ID
- 6778473
-
- 本文言語コード
- ja
-
- データソース種別
-
- JaLC
- NDL
- Crossref
- CiNii Articles
-
- 抄録ライセンスフラグ
- 使用不可
