基于旋翼無人機的農(nóng)業(yè)低空高光譜遙感技術(shù)

高輝 王秋平 趙新剛 潘從元

摘 要：現(xiàn)代農(nóng)業(yè)的發(fā)展要求對農(nóng)作物進行精準管理，需要對大面積農(nóng)作物光譜信息進行精確和實時快速的監(jiān)測，以便大面積、快速、無損的獲取作物的生長、發(fā)育、養(yǎng)分和水分等的空間分布信息，為變量施肥、澆灌、施藥等田間管理提供基本信息，是實現(xiàn)數(shù)字化農(nóng)業(yè)的基礎(chǔ)。利用反射光譜探測作物信息，建立基于特定目標測試數(shù)據(jù)的統(tǒng)計模型來實現(xiàn)作物葉面積指數(shù)、氮素、水分含量和災(zāi)害的光譜反演，已經(jīng)取得了一定的研究成果。但由于建立的統(tǒng)計模型主要是通過地物光譜信息和衛(wèi)星遙感影像數(shù)據(jù)之間的結(jié)合，模型精度受到特定的試驗條件（時間、地點、對象）、空間分辨率、圖譜的特異性等限制，很難進一步提高。該報告提出利用無人機和成像光譜儀結(jié)合，研制出基于無人機的農(nóng)業(yè)低空高光譜的新型遙感技術(shù)平臺，可以靈活地獲取到高分辨率圖譜合一數(shù)據(jù)。結(jié)合衛(wèi)星遙感，將會對不同區(qū)域、不同品種作物的監(jiān)測更靈活、更精確。可以滿足現(xiàn)代農(nóng)業(yè)生產(chǎn)對高空間分辨率、高時間分辨率、高光譜分辨率遙感信息的需求，將為解決精準農(nóng)業(yè)中信息獲取中的瓶頸問題提供一種解決方案。通過該研究實施，在國內(nèi)首次開展了基于無人機的低空農(nóng)作物成像光譜遙感平臺的研制和應(yīng)用示范，為低成本、高效、靈活實時獲取大面積、高空間分辨率的農(nóng)作物信息奠定了一定基礎(chǔ)。報告提出和發(fā)展優(yōu)化了田間農(nóng)作物遙感成像信息探測的多項關(guān)鍵技術(shù)。（1）旋翼無人機超視距自主飛行的控制技術(shù)，包括自主起飛/降落、自主航跡跟蹤、自主容錯、自主定位與優(yōu)化導(dǎo)航的功能；（2）低空遙感應(yīng)用載荷的高精度支撐技術(shù)，用于支撐成像光譜儀系統(tǒng)以獲得高清晰光譜圖像；（3）成像光譜儀的微型化技術(shù)及由此帶來的光譜圖像畸變的矯正方法，用于微型成像光譜儀；（4）提出并實施了無人機懸停狀態(tài)時通過反射鏡旋轉(zhuǎn)，實現(xiàn)光譜數(shù)據(jù)的采集和矯正；（5）建立基于低空高光譜遙感信息的應(yīng)用模型，得到農(nóng)作物不同生育期的成像光譜儀遙感反演模型和相應(yīng)災(zāi)害遙感反演模型，為精準農(nóng)業(yè)的實際應(yīng)用提供依據(jù)。同時還開發(fā)了計算機軟件進行快速數(shù)據(jù)獲取和存儲，并可在短時間內(nèi)完成測得光譜的拼接和數(shù)據(jù)處理。 本報告研究成果還可以應(yīng)用于環(huán)境、食品等科研和企業(yè)單位，可以為農(nóng)業(yè)科研、生產(chǎn)管理、農(nóng)產(chǎn)品質(zhì)量檢測、食品檢測及環(huán)境生態(tài)檢測等提供更好的技術(shù)手段。

關(guān)鍵詞：旋翼無人機 成像光譜儀 遙感 農(nóng)業(yè) 光譜分析

Abstract：Modern agriculture needs the crop spectral information in a large area in time. The spatial resolved spectral information will give the status of crop growth， nutrient and water， which is the basic information for field management as fertilization， irrigation and spraying.So imaging spectral information is the base of digital agriculture. To exploit the spectral information， a spectral information processing model is established that is based on corps reflect spectrum， to find out the leaf area index， nitrogen， moisture and disease Several research results are obtained based on the model. At present， above data processing model is based on the ground spectral information and satellite remote-sensing spectral information， the accuracy of the model is limited by the specific experiment conditions （different time， different height and different spatial resolution）. To improve the accuracy of the data model. The low attitude hyperspectal remote sensing platform based on unmanned aerial vehicle （UAV） is proposed. The platform employs the unmanned aerial vehicle and the image spectrometer， which can be used to get high spatial resolution image and high spectral resolution information for a large area.Combined the UAV based hyperspectral information and the satellite based hyperspectral information， we can monitor crops information in different spatial resolution and flexible way. This platform may be practical used as the high spatial resolution， high spectral resolution spectral information tools for modern agriculture. This is one of the solutions for the bottleneck of the accurate agricultural information collection. In this project， an UAV based hyperspectral information collection platform is developed and evaluated. The platform may provide a low-cost， efficient， flexible and real-time hyperspectral information tool with high spatial resolution of crops spectral information. Several key techniques related to the UAV based hyperspectral sensing for field crops are presented.（1）the UAV control technologies（including UAV take-off and landing， flight track plan， independent fault-tolerant， self-positioning and optimized navigation functions）.（2）the high precision supporting platform for spectrometer.（3）the micro hyperspectrometer development.（4）keeping the hyperspectrometer in a fixed position in the sky， to get hypespectral information by scanning a mirror.（5）developed an application model which may be used on UAV based hyperspectral information. Above research may provide the basis for the practical application of precision agriculture. Sixth， fast hyperspectral data acquisition and storage. The results of this project can also be applied in other research fields as environmental monitoring and food industry.

Key Words：Unmanned Aerial Vehicles；Imaging Spectrometer；Remote Sensing；Agriculture；Spectral Analysis

閱讀全文鏈接（需實名注冊）：http：//www.nstrs.cn/xiangxiBG.aspx？id=39142&flag=1