- 期刊
甘藍水分潛勢基於光譜分析應用於種苗缺水逆境預測之研究
Study on Water Potential of Cabbage Based on Spectral Analysis Applied to Seedling Water Deficiency Stress Prediction
摘要
早期農民未考量土壤與環境變異，導致各定點農作物受到差別影響，長期下來農作物的生產參差不齊，造成農地肥力不均問題。隨著農業體質的轉變和科技的進展，目前使用非破壞、非侵入性的檢測方法成為生物材料測定方法的新趨勢，而光學方法正符合這樣的性質，它能依據不同的檢測目的來進行系統的設計，檢測範圍包含樣本外部與內部品質。本研究從光學分析的角度，以高光譜影像相機檢測波段範圍900～1700nm，並計算不同毫升數供水組別的水潛勢值，研究共使用162片甘藍苗葉片樣本參與建模分析，並採用偏最小平方迴歸法（Partial Least Square Regression）來建立檢量線，其分析結果為決定係數rc2 = 0.642，標準校正誤差為SEC=0.068。本研究延伸探討溫室中種苗生長的檢測與管理，分析何者澆水量為最適值。在應用上，我們將建立一套以Python為基礎的使用者介面，將收集數據導入，建立非破壞性之甘藍種苗缺水逆境預測模型，當光譜相機偵測到缺水逆境情形，將搜尋甘藍苗缺水資料庫後，提供溫室內可變率噴灑系統灌溉之依據。
並列摘要
Early farmers did not consider soil and environmental variation, resulting in differences in crops at designated points. In the long run, resulting in uneven fertility of farmland. With the transformation of agricultural constitution and the advancement of science and technology, the use of non-destructive and non-invasive detection methods has become a new trend of biological material measurement methods, which can be designed according to different detection purposes and the detection range includes the external and internal quality of the sample. In this study, from the perspective of optical analysis, a hyperspectral image camera was used to detect the band range of 900~1700 nm, and the water potential value on each pixel of the leaf image was calculated, a total of 162 leaf samples of cabbage seedlings were used to participate in the modeling analysis. The PLSR (Partial Least Square Regression) prediction model was used to establish the analysis result were rc2 (coefficient of determination of calibration) = 0.642 and SEC (Standard Error of Calibration) = 0.068.This study extends the detection and management of seedling growth in greenhouses, analyzes which watering amount is the most appropriate value. In terms of application, we will establish a set of Python-based user interface, import the collected data, establish a non-destructive cabbage seedling water shortage stress prediction model, when the multispectral camera detects the water shortage adversity, will search the water shortage database of cabbage seedlings, and provide the basis for irrigation of the variable rate spraying system in the greenhouse.