Review
Machine learning approaches for crop yield prediction and nitrogen status estimation in precision agriculture: A review

https://doi.org/10.1016/j.compag.2018.05.012Get rights and content

Highlights

  • Review of studies in crop yield prediction and N status estimation via ML techniques.

  • Comparison of different ML techniques in application to the same task in PA.

  • Discussion on ML techniques used in the reviewed studies.

Abstract

Accurate yield estimation and optimised nitrogen management is essential in agriculture. Remote sensing (RS) systems are being more widely used in building decision support tools for contemporary farming systems to improve yield production and nitrogen management while reducing operating costs and environmental impact. However, RS based approaches require processing of enormous amounts of remotely sensed data from different platforms and, therefore, greater attention is currently being devoted to machine learning (ML) methods. This is due to the capability of machine learning based systems to process a large number of inputs and handle non-linear tasks. This paper discusses research developments conducted within the last 15 years on machine learning based techniques for accurate crop yield prediction and nitrogen status estimation. The paper concludes that the rapid advances in sensing technologies and ML techniques will provide cost-effective and comprehensive solutions for better crop and environment state estimation and decision making. More targeted application of the sensor platforms and ML techniques, the fusion of different sensor modalities and expert knowledge, and the development of hybrid systems combining different ML and signal processing techniques are all likely to be part of precision agriculture (PA) in the near future.

Introduction

Improving crop yield production and quality while reducing operating costs and environmental pollution is a key goal in precision agriculture (PA). The potential growth and yield depends on many different production attributes such as the weather, soil properties, topography, irrigation and fertilizer management. The need for timely and accurate sensing of these inputs for large agricultural fields has led to increased adoption of remote and proximal sensing technologies (Campbell and Wynne, 2011) in PA (Curran, 1987). These sensing techniques provide acquisition of spectral, spatial and temporal information about the objects via ground-based vehicles, aircraft, satellites and handheld radiometers.

Remote sensing, such as satellite and airborne multi-spectral scanning, photography and video, enables precision weed management through the generation of timely and accurate weed maps (Lamb and Brown, 2001). Thermal remote sensing via airborne thermal imagery has the potential to identify spatial variations in crop water status (Tilling et al., 2006), which can enable improvements in the water management in irrigated cropping systems. Nowadays, this remote sensing technique is widely used in different crop species such as wheat (Gontia and Tiwari, 2008), maize (Taghvaeian et al., 2012), trees (Bellvert et al., 2016) and vineyards (Gutiérrez et al., 2018). At the same time many ground-based platforms have been developed and aimed at different PA tasks such as mapping of soil properties (Barnes et al., 2003), estimating evapotranspiration and drought stress (Maes and Steppe, 2012), weed mapping (Sui et al., 2008) and assessing crop water and nitrogen status (El-Shikha et al., 2007, Govender et al., 2009).

Remote sensing at visible and near-infrared wavelengths (vis-NIR) has been used to devise many spectral indices for estimating different vegetation properties. This includes the amount of chlorophylls and other photosynthetic/photoprotective pigments and the leaf area index (LAI) (Barati et al., 2011, Gao, 1996, Haboudane et al., 2004, Huete, 1988, Qi et al., 1994, Sims and Gamon, 2002, Viña et al., 2011, Zarco-Tejada et al., 2012). More than 100 vegetation indices along with their applicability, representativeness, environment and implementation precision have recently been reviewed by Xue and Su (2017). They concluded that for real-world applications the use of any existing vegetation indices requires careful consideration of the strengths and shortcomings of those indices and the specific environment where they will be applied. Making crop yield predictions using remotely sensed vegetation indices has been attempted by Panda et al., 2010, Jaafar and Ahmad, 2015 and many others.

Utilization of wireless sensors and actuators in PA, as well as algorithms for wireless sensor network data integration is now advancing (Zhou et al., 2012). Aqeel-ur-Rehman et al. (2014) presented a review on wireless sensor network technology and their applications in different aspects of agriculture and reported on existing system frameworks in PA.

Nitrogen (N) is considered by growers as a major mineral nutrient for plant growth and development because it is directly related to the photosynthetic process (Andrews et al., 2013). At the same time N has a high environmental and economic impact. Hence, the optimization of N fertilization for different crops has become a subject of many spectrometric studies (Cao et al., 2017, Chen et al., 2008, Goron et al., 2017, Lukina et al., 2001, MacKerron et al., 1993, Maresma et al., 2016, Quemada et al., 2014, Raun et al., 2005, Schepers and Raun, 2008).

The estimation of the plant N status can be divided into two main types: destructive and non-destructive. The most common method of destructive measurement is a chemical analysis which is associated with the Kjeldahl technique and is laborious, lengthy and costly (Jones and Moseley, 1993, Vigneau et al., 2011). Optical remote sensing of the plant N status is a non-destructive method based on canopy reflectance in the visible–NIR wavelengths (400–900 nm). This measurement is completed in-situ, lowering the number of field samples required and thus reducing the time and financial cost of field sample collection, preparation and laboratory analysis. Many studies have been dedicated to non-destructive measurement of the N status inferring in plants via remote sensing technology (Apostol et al., 2007, Lamb et al., 2002, Reyniers and Vrindts, 2006, Scharf et al., 2002, Tremblay et al., 2012) and spectral indices indicative of the plant N status have been derived from hyperspectral data (Chen et al., 2010, Tian et al., 2014, Yao et al., 2010).

Adoption of remote sensing in geology (Gupta, 2003), forestry (Holmgren and Thuresson, 1998, Hultquist et al., 2014), hydrology (Engman and Gurney, 1991), agriculture (Seelan et al., 2003) and other domains has led to the collection of significant volumes of data. The volume is continuously growing and it is beyond human ability to personally integrate, analyse and make the best informed decisions from the information. This is particularly the case when the data is not homogeneous, i.e. is sensed by sensors with different spatial, temporal and spectral modalities. Machine Learning (ML) is an emerging technology that can aid in the discovery of rules and patterns in large sets of data (Du and Jeffrey, 2007).

Crop yield prediction and N status estimation are considered together here because of the direct linkage in fertiliser management decisions. Crop yield goals are routinely utilised for calculating N requirements, both pre- and in-season. For devising potential site-specific management plans for N fertiliser, especially in-season, an estimation of both would be ideal. The aim of this review is to show the capability of different ML techniques to effectively handle these different but closely related tasks. A review is presented of recent studies in the area of crop yield prediction and N status estimation, which incorporate different ML techniques. It also covers comparative studies of different ML techniques as they are applied to the same task in PA. Some technical aspects of the ML techniques used in the reviewed studies are discussed.

Section snippets

Machine learning techniques

One of the main advantages of ML techniques is that they are capable of autonomously solving large non-linear problems using datasets from multiple (potentially interconnected) sources. Some ML techniques, such as Gaussian Processes (GPs) (Bishop, 2006, Rasmussen and Williams, 2005), Dirichlet Processes (DP) (Ferguson, 1973) and Indian Buffet Process (IBP) (Griffiths and Ghahramani, 2011) are probabilistic and enable consideration of sensor noise while conducting probabilistic fusion of

Yield estimation

Achieving maximum crop yield at minimum cost with a healthy ecosystem is one of the main goals of agricultural production. Early detection and management of problems associated with crop yield restrictions can help increase yield and subsequent profit, and estimating yield is important to numerous crop management and business decisions.

In recent years different ML techniques have been implemented to achieve accurate yield prediction for different crops (Subhadra et al., 2016). The most

Precision nitrogen management

As nitrogen (N) plays a significant role in the process of photosynthesis, it is important for crop health and development. At the same time, environmental factors and cost require a prudent application of N. Because of these factors the problem of optimal N management has attracted the attention of numerous researchers (Cao et al., 2015, Cao et al., 2017, Dai et al., 2013, Magney et al., 2017, Zhao et al., 2013).

One of the approaches to optimal N management in PA is to use management zones,

Conclusions

Sensing technologies and ML techniques have rapidly advanced during the last decade. These developments are likely to continue providing cost-effective and more comprehensive datasets combined with more sophisticated algorithmic solutions enabling better crop and environment state estimation and decision making. We are at the beginning of a promising path that has the potential to significantly alter crop yield management.

A number of ML techniques have already been successfully applied to

Acknowledgement

This work has been supported by the Australian Centre for Field Robotics.

References (141)

  • B. Gao

    NDWI—A normalized difference water index for remote sensing of vegetation liquid water from space

    Remote Sens. Environ.

    (1996)
  • N.K. Gontia et al.

    Development of crop water stress index of wheat crop for scheduling irrigation using infrared thermometry

    Agric. Water Manage.

    (2008)
  • Y.L. Grossman et al.

    Critique of stepwise multiple linear regression for the extraction of leaf biochemistry information from leaf reflectance data

    Remote Sens. Environ.

    (1996)
  • P. Guo et al.

    Digital mapping of soil organic matter for rubber plantation at regional scale: An application of random forest plus residuals kriging approach

    Geoderma

    (2015)
  • D. Haboudane et al.

    Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop canopies: Modeling and validation in the context of precision agriculture

    Remote Sens. Environ.

    (2004)
  • P.M. Hansen et al.

    Reflectance measurement of canopy biomass and nitrogen status in wheat crops using normalized difference vegetation indices and partial least squares regression

    Remote Sens. Environ.

    (2003)
  • Z. Huang et al.

    Estimating foliage nitrogen concentration from HYMAP data using continuum removal analysis

    Remote Sens. Environ.

    (2004)
  • A.R. Huete

    A soil-adjusted vegetation index (SAVI)

    Remote Sens. Environ.

    (1988)
  • D.M. Johnson

    An assessment of pre- and within-season remotely sensed variables for forecasting corn and soybean yields in the United States

    Remote Sens. Environ.

    (2014)
  • M. Kaul et al.

    Artificial neural networks for corn and soybean yield prediction

    Agric. Syst.

    (2005)
  • E.B. Knipling

    Physical and physiological bases for the reference of visible and near infrared radiation from vegetation

    Remote Sens. Environ.

    (1970)
  • R.F. Kokaly

    Investigating a physical basis for spectroscopic estimates of leaf nitrogen concentration

    Remote Sens. Environ.

    (2001)
  • D.W. Lamb et al.

    PA—Precision agriculture

    J. Agric. Eng. Res.

    (2001)
  • L. Liang et al.

    Estimation of crop LAI using hyperspectral vegetation indices and a hybrid inversion method

    Remote Sens. Environ.

    (2015)
  • A. Morellos et al.

    Machine learning based prediction of soil total nitrogen, organic carbon and moisture content by using VIS-NIR spectroscopy

    Biosyst. Eng.

    (2016)
  • S. Nawar et al.

    Chapter Four - Delineation of Soil Management Zones for Variable-Rate Fertilization: A Review

  • X.E. Pantazi et al.

    Wheat yield prediction using machine learning and advanced sensing techniques

    Comput. Electr. Agric.

    (2016)
  • E.I. Papageorgiou et al.

    Fuzzy cognitive map based approach for predicting yield in cotton crop production as a basis for decision support system in precision agriculture application

    Appl. Soft Comput.

    (2011)
  • E.I. Papageorgiou et al.

    Yield prediction in apples using Fuzzy Cognitive Map learning approach

    Comput. Electr. Agric.

    (2013)
  • J. Qi et al.

    A modified soil adjusted vegetation index

    Remote Sens. Environ.

    (1994)
  • O. Satir et al.

    Crop yield prediction under soil salinity using satellite derived vegetation indices

    Field Crops Res.

    (2016)
  • I. Ali et al.

    Review of Machine Learning Approaches for Biomass and Soil Moisture Retrievals from Remote Sensing Data

    Remote Sens.

    (2015)
  • M. Andrews et al.

    Do plants need nitrate? The mechanisms by which nitrogen form affects plants

    Ann. Appl. Biol.

    (2013)
  • S. Apostol et al.

    A comparison of multiwavelength laser-induced fluorescence parameters for the remote sensing of nitrogen stress in field-cultivated corn

    Can. J. Remote Sens.

    (2007)
  • E.M. Barnes et al.

    Remote- and ground-based sensor techniques to map soil properties

    Photogramm. Eng. Remote Sens.

    (2003)
  • J. Behmann et al.

    A review of advanced machine learning methods for the detection of biotic stress in precision crop protection

    Prec. Agric.

    (2015)
  • J. Bellvert et al.

    Airborne thermal imagery to detect the seasonal evolution of crop water status in peach, nectarine and saturn peach orchards

    Remote Sens.

    (2016)
  • G.S. Birth et al.

    Measuring the color of growing turf with a reflectance spectrophotometer

    Agron. J.

    (1968)
  • C.M. Bishop

    Pattern Recognition and Machine Learning (Information Science and Statistics)

    (2006)
  • K.L. Bolster et al.

    Determination of carbon fraction and nitrogen concentration in tree foliage by near infrared reflectances: a comparison of statistical methods

    Can. J. For. Res.

    (1996)
  • P. Bose et al.

    Spiking neural networks for crop yield estimation based on spatiotemporal analysis of image time series

    IEEE Trans. Geosci. Remote Sens.

    (2016)
  • L. Breiman

    Random forests

    Mach. Learn.

    (2001)
  • Campbell, J.B., Wynne, R.H., 2011. Introduction to Remote Sensing. fifth ed. Guildford...
  • Q. Cao et al.

    Developing a new crop circle active canopy sensor-based precision nitrogen management strategy for winter wheat in North China Plain

    Prec. Agric.

    (2017)
  • F. Castaldi et al.

    A data fusion and spatial data analysis approach for the estimation of wheat grain nitrogen uptake from satellite data

    Int. J. Remote Sens.

    (2016)
  • D. Chen et al.

    Prospects of improving efficiency of fertiliser nitrogen in Australian agriculture: a review of enhanced efficiency fertilisers

    Soil Res.

    (2008)
  • C. Cilia et al.

    Nitrogen status assessment for variable rate fertilization in maize through hyperspectral imagery

    Remote Sens.

    (2014)
  • R.N. Clark et al.

    Reflectance spectroscopy: Quantitative analysis techniques for remote sensing applications

    J. Geophys. Res.: Solid Earth

    (1984)
  • P.J. Curran

    Remote sensing in agriculture: an introductory review

    J. Geogr.

    (1987)
  • M. Diacono et al.

    Precision nitrogen management of wheat A review

    Agron. Sustain. Dev.

    (2013)
  • Cited by (794)

    View all citing articles on Scopus
    View full text