Please use this identifier to cite or link to this item:
A Robotic System For Transplanting Phalaenopsis Tissue Culture Plantlets
|關鍵字:||Image processing;影像處理;Robot;Machine vision;Phalaenopsis;Transplanting;Lighting;機器人;機器視覺;蝴蝶蘭;移植;打光||出版社:||農業機械工程學系||摘要:||
This research focuses on the combination of image processing technology and cartesian coordinate robot to build a transplanting system for phalaenopsis tissue culture plantlets. This system includes computer vision, computer communication, automatic control, and barometric device. Visual Basic and Matrox Image Library were utilized to develop image-processing system. Phalaenopsis plantlets can be automatically analyzed by computer vision/image system and transplanted by robotic system. Experimental results showed that 78.1 to 86.6% of plantlets were transplanted successfully. The hand efforts of robot move plantlets to the specified position in the tray with an offset less than 2.06mm. If plantlets of transplanted failure were to do again, the 93.89% of plantlets were transplanted successfully in all. In this study, cycle of average was 16.48 second with a plantlets. The absent- plantlets rate was 3.09 with a tray. In efficiency of all was 0.68. The 81.26% of plantlets were transplanted successfully in absent- plantlets sensation and non-recover. If plantlets of transplanted failure were to do again, the 98.85% of plantlets were transplanted successfully in all. In this study, cycle of average was 18.27 second with a plantlets. The absent- plantlets rate was 0 with a tray. In efficiency of all was 0.94. The ratios of success in transplantation are between 78.1% and 86.6%. The reason of failure in transplantation always relates Phalaenopsis plantlets positon or length. Fiber senior were utilized to check effect of holding-plantlets in transplantation. This device was avoidable absent- plantlets and reduced capital.
本研究是利用直角座標機器人結合影像處理設備與技術，建立一套組織培養苗的移植系統，試用於蝴蝶蘭實生苗之分瓶假植作業，以降低人工需求。直角座標機器人藉著電腦通訊、自動控制及氣壓輔助裝置及較無外界光源干擾的底部打光的設計之下，以自動化流程取得幼苗的數位影像，再以自行設計之影像處理系統分析出夾持點與夾持角度，透過機器人的控制，完成整個假植作業。影像處理系統以Visual Basic程式語言配合Matrox Image Library所寫成，能完全自動化進行蝴蝶蘭幼苗影像分析及假植作業，在假植作業的過程中，不需加入人為的判斷及作業。在自動化的移植作業中，扦插位置的平均偏移量為2.06mm。移植成功比例為78.1﹪至86.6﹪之間；移植失敗的原因是蝴蝶蘭種苗在輸送帶上的位置是否理想，以及種苗長度應介於16mm至25mm之間。利用光纖感測做為夾苗效果檢查設備可完全免除移植缺株的問題，有助於降低生產成本，提升電腦機器人系統及套裝程式的等級可做增加機器人移植作業的效率。在此實驗中，每一株的平均週期為16.48秒；每盤的缺株率為3.09株；整體效益為0.68。在缺株感測且假植失敗幼苗不回流實驗之平均成功率為81.26%；若將假植失敗幼苗再回流則成功率高達98.85%。在此實驗中，每一株的平均週期為18.27秒；每盤的缺株率為0株；整體效益為0.93。
|Appears in Collections:||生物產業機電工程學系|
Show full item record
TAIR Related Article
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.