Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/93648
標題: 植物阻抗量測應用於蝴蝶蘭病徵檢測之初探
A Basic Study on Symptoms Detection of Phalaenopsis by Using Electrical Impedance Spectroscopy
作者: Guan-De Wu
Ting-Yi Chiu
Tse-Min Chen
吳冠德
邱庭毅
陳澤民
關鍵字: Disease detection system;Noninvasive detection;Physiological reference curve;Botanical impedance measurement;植物病害檢測;非侵入式檢測法;生理參考曲線;植物阻抗量測
出版社: 臺中巿: 國立中興大學農學院
Project: 農林學報, Volume 63, Issue 1, Page(s) 65-73.
摘要: 
建立一非侵入式植物病害檢測系統為本研究之目的。本研究應用量測分析不同頻譜下蝴蝶蘭的阻抗值,經由不同頻譜下所呈現的阻抗趨勢,期待在病徵出現前能有效篩檢出植株是否已感染根腐或軟腐等常見病害。相關研究指出,植物阻抗在不同頻譜下會出現不同散逸現象,當植物在不同的健康狀態下植物之阻抗頻譜亦會隨之改變,所以植物的生理狀態應可以植物的阻抗值為指標進行檢測。因此若將植物本體視為一包含輸入和輸出的系統,當輸入信號不變,植物內部因結構組成改變時,輸出信號亦會隨之變化;因此輸入一序列不同頻率信號進入植株本體,量測所求得之頻率對應阻抗的關係可做為植物生理的參考曲線。本研究將阻抗量測值以最小平方法建立一轉移函數模型方程組,並利用殘差運算的方式與健康模型方程式進行比較,藉此區辨蝴蝶蘭健康植株與染病植株。當3.5吋蝴蝶蘭阻抗殘差模型讀值超過50或奈奎氏殘差模型讀值超過5時,系統即判斷此植株為染病狀態,實測結果亦顯示在阻抗頻譜殘差的模型圖和奈奎氏殘差模型中皆能有效進行蝴蝶蘭染病預警與判別。

This research established a noninvasive disease detection system for Phalaenopsis. Since the input signals with distinct frequency result in different impedance spectrum as responses, the electrical impedance spectroscopy is an available index to describe the physiological status of botanic tissue. Therefore, employing the measured impedance trend to identify the disease-infected seedlings may be a prospective technique. Former literatures indicated that the electrical impedance of plants changed depended on different frequency spectroscopy and different physiological status, which is called desperation, Desperation depicts the characters of a specific species. If a botanic tissue such as Phalaenopsis is regarded as a system that contains input signal and response and when the input signal is fixed and the plant's structures have been changed, the response will change simultaneously. Thus the frequency spectroscopy of the impedance response of a plant illustrates the biological statues and may be referred as the physiological reference curve. This thesis employs the concept of the referred physiological reference curve of Phalaenopsis and the least square method to organize a novel noninvasive disease detection scheme for Phalaenopsis seedlings. Using the residuals of data from differen experiment days and their fitting transfer models, the Bold and Nyquist analysis validate the effectiveness of this detection system for Phalaenopsis. The limitations of check point value for the residual Bold model is 50 and for Nyquist model is 5. Beyond these check points the seedling is awarded to be infected.
URI: http://hdl.handle.net/11455/93648
Appears in Collections:第63卷 第01期

Files in This Item:
File SizeFormat
85465-6.pdf1.08 MBAdobe PDFView/Open
Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.