节点文献

基于光谱信息的植被氮素快速探测仪器研究

Rapid Detectection Instrument of Vegetation Nitrogen Based on Spectral Information

【作者】 孙刚

【导师】 王秀;

【作者基本信息】 中国农业大学 , 农业机械化工程, 2013, 博士

【摘要】 最近几十年来,我国农业取得巨大成就的同时也面临着日益严重的危机,其中,化肥和农药过量投入问题,给生态环境、农产品安全和农业可持续发展都造成了巨大的压力。为了有效遏制这一局面,需要利用快速、无损、低成本的作物营养及长势信息探测设备来指导农业生产,实现按需施肥,增产的同时减少对环境的破坏。氮素是直接反映植被营养状况及代谢的重要组分,也是长势监测、营养诊断乃至品质评价的重要指标。便携式、实时、准确的植被氮素营养监测与诊断仪器是作物氮肥科学管理基运筹调控的必要手段和关键设备之一。本论文以研制植被冠层氮素的快速探测仪器为目标,以不同的技术方案对植被氮素含量快速探测的关键技术进行了研究,完成的主要研究成果和结论如下:(1)研制了以太阳光为光源,以滤光片为分光器件的被动光源式植被冠层氮素探测仪器—一(NRI-0SAVI)一体化氮素测量仪。该仪器以硅光电探测器为探测传感器,通过测量570nm、670nm和800nm三个特征波长处太阳的入射光和植被冠层的反射光,计算出三个特征波长的反射率,最终计算得到NRI和OSAVI植被指数值,可同时测定植被的氮素含量和长势青况。传感器视场角为60°,测量高度可变。试验表明,该被动式氮素测量仪器的测量值与ASD光谱仪的测量数值相比具有很好的相关性。(2)为了克服被动式光谱测量仪器受环境影响大、工作时间短的劣势,采用LED主动光源脉冲发光技术,进行了主动发光的植被氮素探测仪器研究。首先通过LED主动发出670nm和800nm两个特征波长的脉冲光到探测冠层上,然后再探测该主动光的反射能量,并减去作为背景的太阳光反射能量,计算出两个特征波长处的反射率。接着,利用反射率计算得到反映冠层氮素水平的NDVI植被指数值。该传感器在一定范围内测量结果不受探测高度的影响。实验表明,本氮素测量仪器的测量值与ASD光谱仪的测量数值相比具有很好的相关性。在不同光照条件下的测试表明,该仪器的测量值受太阳光环境影响较小。(3)由于氮素“易运转”的特性,植被叶片氮素含量在植被上表现出明显的垂直分布特性,本论文研究引入多波段激光雷达技术,获取植被的完整的回波信息,开展了植被氮素含量垂直探测的试验研究。首先推导了复杂散射体的激光雷达方程,然后用6种波段针对3个不同的场景进行了回波模拟,探讨了多波段激光雷达在植被氮素含量方面的潜力。接着,详细介绍了多光谱全波形激光雷达样机的研制。该激光雷达以超连续谱脉冲激光器为光源,以望远镜为信号收集单元,以高速光电转换器为传感器,具有同轴发送接收,全波形记录的特点。其探测的距离分辨率约为15cm,探测的发散角为4mrad,最佳探测距离为20米。根据分光方式不同,本文提出并试验了两种光电探测方案,其一是利用滤光片分光,配合两个分立的雪崩二极管APD模块,实现2个波段的波形探测:其二是利用光栅分光,配合光电倍增管PMT阵列,实现32通道的波形探测。目前的实验表明,方案一可以获取较好的探测波形,能满足设计要求。而方案二由于PMT线阵输出的值较小,获取的波形不太理想,还需要进一步改进。试验表明,该仪器在测距精度方面优于15cm,在多光谱信息获取方面,信噪比较高,能构造植被指数实现分类,还可实现氮素含量垂直探测。整体而言,所设计的仪器具备了生产具有光谱信息的3D点云能力。

【Abstract】 In recent years, China’s agriculture has made great achievemnets as well as facing increasingly serious crisis. The excessive investment of chemical fertilizers and pesticides gave great pressure to the safety of agricultural products, ecological environment and the sustainalble development of agricuture. In order to effectively curb this situation, we need fast, non-destructive and low cost of crop nutrition and growth information detection equipment to guide the agricultural production to achieve fertilization according to needs, increase production and reduce damage to the environment. Nitrogen is a metabolism of vegetation which directly reflects the mutrition. It is an important index for monitoring of corop growth, nutrition diagnosis and quality evaluation. Portable, real-time,accurate vegetation nitrogen monitoring and diagnosis instrument is one of the necessary means and key equipment for crop nitrogen management. The goal of this thesis is to develop rapid detection instrument for the canopy nitrogen. Three different technologies were studied. The followings are the main research results and conclusions.1) A novel instrument for measuring nitrogen refect index (NRI) and optimization soil-adjusted vegetation index (OSAV1) is introduced. Through measuring the incident light and the reflected light at570nm,670nm and800nm, the instrument could measure the vegetation index accurately according to the automatic calibration techniques based on the least squares estimation. By using the signal processing method, the problem caused by the weak signal and the sun angle was solved. This instrument measured the vegetation index conveniently, and the results had a very good correlation with the values measured by ASD.The experiments indicated that instrument could be used for non-destructive measuring the crop nitrogen. The instrument had the characters of low power consumption, high sensitivity and rapid response time. The crop nutrient measured by the instrument could be used for variable rate fertilization.2) In order to overcome the disadvantages of passive measuring instrument, the LED active light pulse light was introduced and the active vegetation nitrogen detection instrument was researched. Firstly, the light with wavelengths at670nm and800nm was emitted by two kinds of LED and injected to the target. And the refection light was detected and the sunlight was subtracted as background. The the reflection and NDVI was calculated.The field of view of the instrument is60°, and the measuring results are not affected by the probe height. Experiments show that the result of the instrument has good correlation with the result tested by the ASD spectrometer. The environmental impact to the instrument is very little.3) Because of nitrogen is the elements which are easy transfer, nitrogen vertical distribution is common on many vegetation. Considering the retrieval limitation relying only on solar illumination, the multispectral full-wave lidar will be studied for the retrieval of nitrogen concentration vertical distribution. Firstly, solid scatterer model according to the back scattering feature of vegetaion objects was established. Secondly, the multi-channel LiDAR waveform was simulated with the consideration of single scatter of plant radiative transfer model. The variation of LAI and chlorophyll along plant height was investigated for maize and three typical scenarios were assumed. The multi-channel return full waveform LiDAR simulation was carried out based on reflectance calculated by radiative transfer model. The simulation at specific wavelength (531nm,550nm,670nm,780nm,1064nm,1550nm) showed that the multi-channel LiDAR waveform could reflect the variation of plant physiological composition and facilitate the inversion algorithm design. At last, the multi-channel LiDAR prototype was developed. The light source used is a supercontinuum laser source with an average output power of100mW. A telescope collected the scattered laser pulse from the target. According to the different of the sensor and spectrophotometric method, we proposed two schemes. In the first scheme, the focal point of the telescope is imaged onto two avalanche photodiode sensors. And we can get two-channel LiDAR. In the second scheme, the focal point of the telescope is transformed to a grating, and the splitled light was measured by the PMT line sensor. A32-channel LiDAR was achieved. Expeiment shows that the firt scheme can obtain better waveform. Using this instrument,3D clouds with spectral information will be got.

【关键词】 植被氮素快速探测探测仪器反射光谱激光雷达
【Key words】 vegetationnitrogenrapaid detectioninstrumentspectrume
节点文献中: 

本文链接的文献网络图示:

本文的引文网络