论文部分内容阅读
利用植物和背景(枯枝、土壤等)的光谱特性“红边”两侧反射率的差异,研究了探测绿色植物靶标的光谱探测技术。定义850与650nm处反射率的比值为植物判别指数(GPDI)。用FieldSpec Handheld2500型野外便携式光谱仪测量了绿色植物和背景的光谱数据,对其进行数据处理,计算各被测物质的植物判别指数GPDI。利用决策树模式识别方法建立植物与背景的分类模型,得到了GPDI阈值(GPDITH),选择此阈值为5.54。当GPDI>GPDITH时,判别探测对象为植物;反之亦然。设计开发了基于AT89S51单片机和光电二极管OPT101的绿色植物光谱探测器。试验结果表明,此探测器的探测率受杂草的种类、大小和密度的影响;阔叶草比窄叶草更易探测到;植株越大、密度越高,探测率越高。
Using spectral characteristics of plant and background (litter, soil, etc.) and differences in reflectivity on both sides of “red edge”, spectral detection techniques for detecting green plant targets were studied. The ratio of reflectance at 850 and 650 nm is defined as the plant discrimination index (GPDI). The spectral data of green plants and background were measured with FieldSpec Handheld2500 field portable spectrometer. The data were processed and the plant discrimination index (GPDI) of each tested substance was calculated. The classification model of plant and background was established by using the decision tree pattern recognition method. The GPDI threshold was obtained, and the threshold was selected as 5.54. When GPDI> GPDITH, the detection target is a plant; vice versa. Design and development of a green plant spectrum detector based on AT89S51 microcontroller and photodiode OPT101. The test results show that the detection rate of this detector is affected by the type, size and density of weeds; broadleaf grass is more easily detected than narrowleaf grass; the larger the plant, the higher the density, the higher the detection rate.