Journal of Integrative Agriculture ›› 2021, Vol. 20 ›› Issue (11): 2880-2891.DOI: 10.1016/S2095-3119(20)63556-0

所属专题: 麦类耕作栽培合辑Triticeae Crops Physiology · Biochemistry · Cultivation · Tillage

• 论文 • 上一篇    下一篇

  

  • 收稿日期:2020-03-24 出版日期:2021-11-01 发布日期:2021-09-17

Estimating wheat fractional vegetation cover using a density peak k-means algorithm based on hyperspectral image data

LIU Da-zhong, YANG Fei-fei, LIU Sheng-ping   

  1. Key Laboratory of Agricultural Information Service Technology, Ministry of Agriculture and Rural Affairs/Intelligent Agriculture Research Office, Agricultural Information Institute, Chinese Academy of Agricultural Sciences, Beijing 100081, P.R.China
  • Received:2020-03-24 Online:2021-11-01 Published:2021-09-17
  • Contact: Correspondence LIU Sheng-ping, E-mail: liushengping@caas.cn
  • Supported by:
    This work was supported by the Beijing Natural Science Foundation, China (4202066), the Central Public-interest Scientific Institution Basal Research Fund, China (JBYW-AII-2020-29 and JBYW-AII-2020-31), the Key Research and Development Program of Hebei Province, China (19227407D), and the Technology Innovation Project Fund of Chinese Academy of Agricultural Sciences (CAAS-ASTIP-2020-All).

摘要:

植被覆盖度(FVC)是衡量作物生长状况的重要指标。在作物生长监测研究中,快速、准确地提取植被覆盖度非常重要。摄影法是目前应用最广泛的FVC提取方法,具有操作简单、提取精度高的优点。但是,当土壤湿度和采集时间不同时,提取结果的准确度较差。为了适应不同的植被覆盖度提取条件,本文提出了一种新的植被覆盖度提取方法,该方法采用密度峰值K-means (density peak K-means, DPK-means)算法从小麦归一化差值植被指数(NDVI)灰度图像中提取植被覆盖度。本研究以盆栽种植的杨麦4 (YF4)和田间种植的杨麦16 (Y16)为研究对象,使用三脚架搭载高光谱成像相机,在盆栽小麦冠层上方1m处采集不同土壤条件(干、湿)下小麦的地面高光谱图像。无人机搭载高光谱相机,在田间小麦冠层上方50m高空采集不同时期的冬小麦高光谱图像。分别采用像元二分法和DPK-means算法对小麦NDVI灰度图像中的植被像元和非植被像元进行分类,并对两种方法的提取效果进行了比较分析。结果表明,像素二分法提取的图像受到采集条件的影响较大,误差分布较为分散。DPK-means算法的提取效果受采集条件的影响较小,误差分布比较集中。干、湿土壤条件和不同时间条件下的误差绝对值分别为0.042和0.044,均方根误差(RMSE)分别为0.028和0.030,FVC拟合精度R2分别为0.87和0.93。本研究表明,在不同土壤和时间条件下,DPK-means算法比像元二分法能够获得更准确的结果,是一种准确、稳健的FVC提取方法。


Abstract:

Fractional vegetation cover (FVC) is an important parameter to measure crop growth.  In studies of crop growth monitoring, it is very important to extract FVC quickly and accurately.  As the most widely used FVC extraction method, the photographic method has the advantages of simple operation and high extraction accuracy.  However, when soil moisture and acquisition times vary, the extraction results are less accurate.  To accommodate various conditions of FVC extraction, this study proposes a new FVC extraction method that extracts FVC from a normalized difference vegetation index (NDVI) greyscale image of wheat by using a density peak k-means (DPK-means) algorithm.  In this study, Yangfumai 4 (YF4) planted in pots and Yangmai 16 (Y16) planted in the field were used as the research materials.  With a hyperspectral imaging camera mounted on a tripod, ground hyperspectral images of winter wheat under different soil conditions (dry and wet) were collected at 1 m above the potted wheat canopy.  Unmanned aerial vehicle (UAV) hyperspectral images of winter wheat at various stages were collected at 50 m above the field wheat canopy by a UAV equipped with a hyperspectral camera.  The pixel dichotomy method and DPK-means algorithm were used to classify vegetation pixels and non-vegetation pixels in NDVI greyscale images of wheat, and the extraction effects of the two methods were compared and analysed.  The results showed that extraction by pixel dichotomy was influenced by the acquisition conditions and its error distribution was relatively scattered, while the extraction effect of the DPK-means algorithm was less affected by the acquisition conditions and its error distribution was concentrated.  The absolute values of error were 0.042 and 0.044, the root mean square errors (RMSE) were 0.028 and 0.030, and the fitting accuracy R2 of the FVC was 0.87 and 0.93, under dry and wet soil conditions and under various time conditions, respectively.  This study found that the DPK-means algorithm was capable of achieving more accurate results than the pixel dichotomy method in various soil and time conditions and was an accurate and robust method for FVC extraction. 

Key words: fractional vegetation cover ,  k-means algorithm ,  NDVI , vegetation index ,  wheat