The Study of Intelligent Campus Leaf Species Recognition

• Main Authors: Kuei-Long Huang, 黃奎龍
• Other Authors: Wen-Ping Chen
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/10144516008576529744

碩士 === 國立高雄應用科技大學 === 電機工程系 === 99 === In order to develop an intelligent leaf recognition system for campus plants, this thesis proposes three-stages leaf automatic recognition systems for leaf outside. In the first stage, according to leaf's outside classification, eliminating impossible species. Then, according to the filtered species of the first stage to delete shape and query image that were large differences. Finally, in the third-stage, configure weights to different characteristic values to find the nearest species. I collected 20 campus plants from National Kaohsiung University of Applied Sciences and National Feng-Hsin Senior High School, each kind of plant random sample 25 pieces of leaves. Experimental result indicates that recognition rate of 10 species reach 90% and recognition rate of 20 species is 83%.