Precision Agriculture Research Research Articles

A spraying path planning algorithm based on colour-depth fusion segmentation in peach orchards

Automatic and precise spraying in orchards is currently an active area in precision agriculture research. Research on this topic mainly focuses on the recognition of fruit trees and the planning of driving paths. The dense canopy and complex background of peach tree orchards present great challenges in fruit tree recognition. In this study, an algorithm based on colour-depth vision was developed to solve the path planning problem for precise spraying in peach orchards. First, a colour-depth binocular sensor was used to acquire video images in peach orchards. Then, a colour-depth fusion segmentation method (CDFS) based on the leaf wall area (LWA) of the colour-depth images was proposed. This method combines the result of Otsu segmentation for the green layer of the colour image and the K-means layering result of the depth image to accurately detect the LWA of peach trees. Then, image erosion was used to delimit the two largest LWAs as the region of interest (ROI). Finally, the spraying path was planned by detecting the midpoint of ROI spacing as the end of the spraying path. A comparative analysis of five image segmentation algorithms showed that the proposed CDFS algorithm produced the closest LWA to that obtained by manual segmentation and thus achieved the best segmentation effect. Comparing this automatic path plan method with the drivable areas showed that the overall accuracy rate of the path planning was 97.5%. Therefore, the proposed algorithm planned paths within the drivable area and can thus be used to accurately and automatically plan spraying paths in peach orchards, except when there were gaps between rows in the fruit orchard.

On spline-based approaches to spatial linear regression for geostatistical data

For spatial linear regression, the traditional approach to capture spatial dependence is to use a parametric linear mixed-effects model. Spline surfaces can be used as an alternative to capture spatial variability, giving rise to a semiparametric method that does not require the specification of a parametric covariance structure. The spline component in such a semiparametric method, however, impacts the estimation of the regression coefficients. In this paper, we investigate such an impact in spatial linear regression with spline-based spatial effects. Statistical properties of the regression coefficient estimators are established under the model assumptions of the traditional spatial linear regression. Further, we examine the empirical properties of the regression coefficient estimators under spatial confounding via a simulation study. A data example in precision agriculture research regarding soybean yield in relation to field conditions is presented for illustration.

A Deep Learning Semantic Segmentation-Based Approach for Field-Level Sorghum Panicle Counting

Small unmanned aerial systems (UAS) have emerged as high-throughput platforms for the collection of high-resolution image data over large crop fields to support precision agriculture and plant breeding research. At the same time, the improved efficiency in image capture is leading to massive datasets, which pose analysis challenges in providing needed phenotypic data. To complement these high-throughput platforms, there is an increasing need in crop improvement to develop robust image analysis methods to analyze large amount of image data. Analysis approaches based on deep learning models are currently the most promising and show unparalleled performance in analyzing large image datasets. This study developed and applied an image analysis approach based on a SegNet deep learning semantic segmentation model to estimate sorghum panicles counts, which are critical phenotypic data in sorghum crop improvement, from UAS images over selected sorghum experimental plots. The SegNet model was trained to semantically segment UAS images into sorghum panicles, foliage and the exposed ground using 462, 250 × 250 labeled images, which was then applied to field orthomosaic to generate a field-level semantic segmentation. Individual panicle locations were obtained after post-processing the segmentation output to remove small objects and split merged panicles. A comparison between model panicle count estimates and manually digitized panicle locations in 60 randomly selected plots showed an overall detection accuracy of 94%. A per-plot panicle count comparison also showed high agreement between estimated and reference panicle counts (Spearman correlation ρ = 0.88, mean bias = 0.65). Misclassifications of panicles during the semantic segmentation step and mosaicking errors in the field orthomosaic contributed mainly to panicle detection errors. Overall, the approach based on deep learning semantic segmentation showed good promise and with a larger labeled dataset and extensive hyper-parameter tuning, should provide even more robust and effective characterization of sorghum panicle counts.

Deployment and Verifications of the Spatial Filtering of Data Measured by Field Harvesters and Methods of Their Interpolation: Czech Cereal Fields between 2014 and 2018.

Yield mapping is a subject of research in (precision) agriculture and one of the primary concerns for farmers as it forms the basis of their income and has implications for subsidies and taxes. The presented approach involves deployment of field harvesters equipped with sensors that provide more detailed and spatially localized values than merely a sum of yields for the whole plot. The measurements from such sensors need to be filtered and subject to further processing, including interpolation, to facilitate follow-up interpretation. This paper aims to identify the relative differences between interpolations from (1) (field) measured data, (2) measured data that were globally filtered, and (3) measured data that were globally and locally filtered. All the measured data were obtained at a fully operational farm and are considered to represent a natural experiment. The revealed spatial patterns and recommendations regarding global and local filtering methods are presented at the end of the paper. Time investments into filtering techniques are also taken into account.

UNMANNED AERIAL VEHICLE (UAV) DERIVED NORMALISED DIFFERENCE VEGETATION INDEX (NDVI) AND CROWN PROJECTION AREA (CPA) TO DETECT HEALTH CONDITIONS OF YOUNG OIL PALM TREES FOR PRECISION AGRICULTURE

Abstract. Malaysia currently is one of the biggest global producers and exporters of palm oil. The world’s expanding oil palm plantation areas contribute to climate change and in-return, climate is change also affecting the health of oil palms through a range of abiotic and biotic stresses. Current advancements in Precision Agriculture research using UAV gives an advantage to detect the health conditions of oil palm at early stages. Thus, remedial actions can be taken to prolong the life and increase oil palms productivity. This paper explores the use of UAV derived NDVI and CPA of young oil palm to detect the health conditions. NDVI of individual oil palm were extracted using ground masking layer from the dense point clouds and visual on-screen manual editing was done for removing trees other than oil palm in ENVI software. The classified individual crown NDVI were then processed to extract the mean NDVI also conversion to vector to obtain the individual crown outline. Extracted mean NDVI was classified into un-healthy and healthy trees while the CPA was classified into small, medium and big size classes. These classes of NDVI and CPA were analysed using GIS overlay method thus revealing the spatial patterns of individual oil palm trees and its health conditions. Overall, the majority of oil palm trees of the study area are healthy but average performing. However, few oil palm trees detected having health problems which has low NDVI and small CPA. This study demonstrates that biophysical parameters such as the CPA can be used to detect individual young oil palm trees health conditions and problems when combined with vegetation indices such as NDVI.

World Sulfur Use Efficiency for Cereal Crops

Sulfur (S) is an essential plant nutrient needed for higher crop yields and improved nutritional value. In recent decades, the occurrence of S deficiency has increased and fertilizer S use may steadily increase. This may lead to inefficient crop utilization of S and result into negative footprints on the environment. The objective of this work was to estimate world fertilizer sulfur use efficiency (SUE) for major cereal crops grown around the world. A 10‐yr data set (2005–2014) was obtained from the Food and Agriculture Organization, the US Geological Survey, and an array of other published research articles. Statistical analysis was performed using MS Excel to obtain total area for world and cereal crops, grain yield, and fertilizer S applied. The difference method [(Total grain S – grain S derived from the soil)/S applied] was used to compute world SUE. Cereal crops included in this study were barley (Hordeum vulgare L.), maize (Zea mays L.), rice (Oryza sativa L.), millet (Pennisetum glaucum L.), wheat (Triticum aestivum L.), sorghum (Sorghum bicolor L.), rye (Secale cereale L.), and oat (Avena sativa L.). Cereal production increased from 2669 M Mg in 2005 to 3346 M Mg in 2014. Sulfur use efficiency for cereal crops was estimated to be 18%. This low SUE may be attributable to S leaching from the soil profile, immobilization, retention in residues, and adsorption. As increased quantities of fertilizer S are likely to be applied in future to meet the ever‐growing demand for food, SUE could decline below 18%.Core Ideas World sulfur use efficiency for cereal crops is unknown. World sulfur use efficiency for cereal crops was estimated to be 18%. More precision agriculture research is necessary to improve sulfur use efficiency for cereal crops. Reasons for low sulfur use efficiency include sulfur; leaching, adsorption, retention in residues, and immobilization as well as failure to adhere to sound agronomic practices and 4R concepts.

UAVs technology for the development of GUI based application for precision agriculture and environmental research

Farmers, agencies, agricultural research community and firms require access to tools to analyze and estimate stressed and productive regions to obtain higher yield. At present, this is performed manually using visual interpretation. Recently there has been some development in the detection and mapping of the stressed crop by use of hyperspectral analysis; but, there is an information gap between farmers and information about the location of the crop under stress in the given area. There is an urgent need to provide a robust solution to identify the stressed region in the agricultural area. To address this, a unique application called as VegNet (Vegetative-Network) has been developed, which aims to provide the necessary tools to detect stressed crop locations using the spectral images obtained from UAVs, and provide stressed crops condition, location and area covered by those stressed crops. In this paper, a combination of spectral vegetation indices techniques has been highlighted to produce a comprehensive solution for precision agriculture using a UAV and VegNet. This incorporates several algorithms; segmentation, Canny-edge detector, dilation, gap-filling, image extraction and locating the stressed region using spectral modelling based Graphical User Interface (GUI) application for precision agriculture, societal benefit and Environmental research.

An approach to forecast grain crop yield using multi-layered, multi-farm data sets and machine learning

Many broadacre farmers have a time series of crop yield monitor data for their fields which are often augmented with additional data, such as soil apparent electrical conductivity surveys and soil test results. In addition there are now readily available national and global datasets, such as rainfall and MODIS, which can be used to represent the crop-growing environment. Rather than analysing one field at a time as is typical in precision agriculture research, there is an opportunity to explore the value of combining data over multiple fields/farms and years into one dataset. Using these datasets in conjunction with machine learning approaches allows predictive models of crop yield to be built. In this study, several large farms in Western Australia were used as a case study, and yield monitor data from wheat, barley and canola crops from three different seasons (2013, 2014 and 2015) that covered ~ 11 000 to ~ 17 000 hectares in each year were used. The yield data were processed to a 10 m grid, and for each observation point associated predictor variables in space and time were collated. The data were then aggregated to a 100 m spatial resolution for modelling yield. Random forest models were used to predict crop yield of wheat, barley and canola using this dataset. Three separate models were created based on pre-sowing, mid-season and late-season conditions to explore the changes in the predictive ability of the model as more within-season information became available. These time points also coincide with points in the season when a management decision is made, such as the application of fertiliser. The models were evaluated with cross-validation using both fields and years for data splitting, and this was assessed at the field spatial resolution. Cross-validated results showed the models predicted yield relatively accurately, with a root mean square error of 0.36 to 0.42 t ha−1, and a Lin’s concordance correlation coefficient of 0.89 to 0.92 at the field resolution. The models performed better as the season progressed, largely because more information about within-season data became available (e.g. rainfall). The more years of yield data that were available for a field, the better the predictions were, and future work should use a longer time-series of yield data. The generic nature of this method makes it possible to apply to other agricultural systems where yield monitor data is available. Future work should also explore the integration of more data sources into the models, focus on predicting at finer spatial resolutions within fields, and the possibility of using the yield forecasts to guide management decisions.

Clustering Field-Based Maize Phenotyping of Plant-Height Growth and Canopy Spectral Dynamics Using a UAV Remote-Sensing Approach.

Phenotyping under field environmental conditions is often considered as a bottleneck in crop breeding. Unmanned aerial vehicle high throughput phenotypic platform (UAV-HTPP) mounted with multi-sensors offers an efficiency, non-invasive, flexible and low-cost solution in large-scale breeding programs compared to ground investigation, especially where measurements are time-sensitive. This study was conducted at the research station of the Xiao Tangshan National Precision Agriculture Research Center of China. Using the UAV-HTPP, RGB and multispectral images were acquired during four critical growth stages of maize. We present a method of extracting plant height (PH) at the plot scale using UAV-HTPP based on the spatial structure of the maize canopy. The core steps of this method are segmentation and spatial Kriging interpolation based on multiple neighboring maximum pixels from multiple plants in a plot. Then, the relationships between the PH extracted from imagery collected using UAV-HTPP and the ground truth were examined. We developed a semi-automated pipeline for extracting, analyzing and evaluating multiple phenotypic traits: canopy cover (CC), normalized vegetation index (NDVI), PH, average growth rate of plant height (AGRPH), and contribution rate of plant height (CRPH). For these traits, we identify genotypic differences and analyze and evaluate dynamics and development trends during different maize growth stages. Furthermore, we introduce a time series data clustering analysis method into breeding programs as a tool to obtain a novel representative trait: typical curve. We classified and named nine types of typical curves of these traits based on curve morphological features. We found that typical curves can detect differences in the genetic background of traits. For the best results, the recognition rate of an NDVI typical curve is 59%, far less than the 82.3% of the CRPH typical curve. Our study provides evidence that the PH trait is among the most heritable and the NDVI trait is among the most easily affected by the external environment in maize.

Use of Farmer Knowledge in the Delineation of Potential Management Zones in Precision Agriculture: A Case Study in Maize (Zea mays L.)

One of the fields of research in precision agriculture (PA) is the delineation of potential management zones (PMZs, also known as site-specific management zones, or simply management zones). To delineate PMZs, cluster analysis is the main used and recommended methodology. For cluster analysis, mainly yield maps, remote sensing multispectral indices, apparent soil electrical conductivity (ECa), and topography data are used. Nevertheless, there is still no accepted protocol or guidelines for establishing PMZs, and different solutions exist. In addition, the farmer’s expert knowledge is not usually taken into account in the delineation process. The objective of the present work was to propose a methodology to delineate potential management zones for differential crop management that expresses the productive potential of the soil within a field. The Management Zone Analyst (MZA) software, which implements a fuzzy c-means algorithm, was used to create different alternatives of PMZ that were validated with yield data in a maize (Zea mays L.) field. The farmers’ expert knowledge was then taken into account to improve the resulting PMZs that best fitted to the yield spatial variability pattern. This knowledge was considered highly valuable information that could be also very useful for deciding management actions to be taken to reduce within-field variability.

Public–private collaboration toward research, education and innovation opportunities in precision agriculture

Regular assessment of precision agriculture (PA) knowledge gaps and opportunities, of PA’s relationship to decision and sustainable agriculture, and of the roles of various research and educational institutions helps advance the development, use and adoption of this science. It also helps to limit redundancy in a science that is rapidly evolving. Such assessments were conducted during a 2016 roundtable discussion by 32 public-sector and industry scientists, administrators and stakeholders predominately from the United States but also from Brazil, Canada, Germany, Israel, New Zealand, South Korea, and the United Kingdom. This article synthesizes those discussions and presents perspectives on developing coordinated, useable decisions from PA to meet sustainable agriculture goals. It concludes with milestones needed to advance PA research and development. Results will guide both public and private researchers and other stakeholders seeking to develop PA to help meet national and global sustainability goals for agriculture. These efforts will also act as a catalyst for regular discussions on PA research, education and development opportunities and coordination between public and private institutions in international forums on PA. Expected impacts include global action to help bring PA techniques and technologies into the mainstream of farming systems and acceleration of the interaction of genetics, environment, management and socioeconomic [(G × E × M) S] approach to sustainably intensify agriculture.

Science mapping approach to analyze the research evolution on precision agriculture: world, EU and Italian situation

The relevance of precision agriculture produced a growth of the related literature over the years. However, a structured analysis of the published material is still missing. Thus, this study attempts to analyze the global scientific output of precision agriculture researches published during the period 2000–2016. By using a science mapping approach, mainly based on the application of network analysis tools, it was possible to investigate pivotal aspects of this research field such as publication trends, research topics and their geographical distribution. Using the Scopus database 17,756 scientific publications were retrieved from the chosen period. The number of publications increased after 2006, highlighting the vibrancy of the field. By authoring 35% of the publications, U.S.A. and China were the most active knowledge producer countries. Moreover, the generation of time resolved maps allowed us to identify agriculture engineering, computer science and agriculture studies as three main research areas characterizing precision agriculture panorama. The paper discusses the distribution of these topics at global level, among European countries and in Italy. Overall, this analysis represents the first holistic view of precision agriculture research providing valuable information for farmers, policy makers and researchers.

ASD — Highland Precision Agriculture case study: Improving precision agriculture with visible and near-infrared spectroscopy

Precision agriculture is an innovative and integrated approach to agriculture that aims to improve the efficiency of resource use and the decision-making surrounding variation on farms. During the past few decades, investments in precision agriculture research and technology have significantly increased worldwide. One such technological advance, near infrared spectroscopy, has been used as an analytical tool in the agriculture industry since the 1990s, and as instruments continue to evolve over time, the use of near infrared continues to progress. In this paper, current developments and future applications of visible and near-infrared spectroscopy (Vis-NIR) are discussed. The research aims to determine whether different pest infestation levels (in this case, the two-spotted spider mite) and other plant stressors on strawberry leaves can be distinguished spectrally using high-resolution Vis-NIR. Since research is still underway, the development of commercially viable monitoring methods remains in its infancy. However, the goal is to advance remote sensing techniques and image analytics to the point where fields can be scanned more quickly, efficiently, and accurately than with traditional monitoring methods.

Real-Time Monitoring of Soil Compaction Using Piezoceramic-Based Embeddable Transducers and Wavelet Packet Analysis

Quantitative assessment of soil compaction is of great importance to construction work, precision agriculture, and geotechnical research. Currently, measurement of soil compaction is mainly performed through field measurements, and the accuracy is highly influenced by personnel experience. In addition, most of the current methods for soil compaction measurement are capable of real-time monitoring. This paper presents an exploratory study of using a piezoceramic-based active sensing approach coupled with wavelet packet analysis to quantitatively monitor soil compaction in real time. In the active sensing approach, a pair of smart aggregates (SA) consisting of piezoceramic transducers embedded inside a host material are embedded into soil. In the pair, one SA acts as transmitter to transmit a modulated stress wave towards the other SA, which acts as receiver. The attenuation ratio of the stress wave propagating energy in soil is governed by the degree of soil compaction, so that the received wave properties of the smart aggregate sensor can be further characterized by using wavelet packet as an index to evaluate the soil compaction in real time. Integrated with remote sensing technologies, the proposed method has potential to be utilized as a real-time remote-sensing technology, which can offer a solution in the monitoring the degree of soil compaction.

Architecting an IoT-enabled platform for precision agriculture and ecological monitoring: A case study

This paper discusses a case study of designing a private Internet of Things (IoT) enabled platform for the research in precision agriculture and ecological monitoring domains. The system architecture is gradually derived using an approach of multiple, concurrent views. Each view represents an architectural perspective describing the solution from the viewpoint of different stakeholders, such as end-users, researchers, developers, and project managers. The end-user requirements have been identified using a set of high-level scenarios, which capture the context and illustrate the motivation for building the platform. The requirements and architecture of the proposed platform have been derived so that the users of the platform, researchers, and developers on the project, can utilize it for prototyping solutions for these high level use cases. The paper further describes the implementation of the platform and its evaluation using various sensor nodes deployed at the research and end-user facilities. The solution is open to further development with respect to supporting additional IoT protocols, data types, and interfacing to various analytics tools. The proposed architecture can also be implemented using different server platforms and cloud technologies.

Regional and Rural Broadband Policy and Programming

Higher speed broadband services are considered as fundamental requirement for digital economy. Availability and usability of internet connected applications is now demanding more bandwidth and capacity in broadband services. Rural people are not excluded from this wave of usage. Focusing on SW Ontario, my research project will employ continuously updated datasets and develops and conducts outcome analysis for public broadband investment, primarily with partners in the Southwest Integrated Fibre Technology (SWIFT) network. The research will combine empirical research methods with an underlying action research process recommended by relevant literature that encourages analytical tools and findings for relevant stakeholders. The research will advance existing efforts to conduct collaborative, well-structured and systematically updated quantitative data analysis for reporting on opportunities and obstacles in current and prospective digital economy infrastructure, applications and adoption in SW Ontario. This analysis will inform evidence-based policy and decision making in various governments. This research will also support other agri-food/rural related policy and programming focused on specific digital investments including identifying connectivity for new ICT applications useful for cost-savings in rural public services or 'next generation' agri-food network technologies (e.g. Precision agriculture research). Health and education sector applications are also relevant. This analysis will support job growth and economic benefits in rural Ontario.

Using an inexpensive color sensor for rapid assessment of soil organic carbon

Quantifying soil organic carbon (SOC) is important for soil management, precision agriculture, soil mapping and carbon dynamics research. Inexpensive sensor technologies offer the potential for rapid quantification of SOC in laboratory samples as well as in the field. The objective of this study was to use a commercially-available color sensor to develop SOC prediction models for both dry and moist soils from the Piedmont region of South Carolina. Thirty-one soil samples were analyzed for lightness to darkness, redness to greenness, and yellowness to blueness (CIEL*a*b*) color using a Nix Pro™ color sensor. Soil color was measured under both dry and moist soil conditions and the depth of each soil sample was also recorded. Using L*, a*, b* and soil depth for each sample as initial predictors, regression analyses were conducted to develop SOC prediction models for dry and moist soils. The resulting residual plots, root mean squared errors (RMSE), and coefficients of determination (R2) were used to assess model fits for predicting the SOC content of soil. Cross validation was conducted to determine the efficiency of the predictive models and the mean squared prediction error (MSPE) was calculated. The final models included soil depth, L*, and a* as independent variables (dry soils R2=0.7978 and MSPE=0.0819, moist soils R2=0.7254 and MSPE=0.1536). The results suggest that soil color sensors have potential for rapid SOC determination, and soil depth and color are useful in predicting SOC content in soils.

Rice Growth Monitoring by Means of X-Band Co-polar SAR: Feature Clustering and BBCH Scale

Precision agriculture research, which aims to monitor agricultural fields and to manage agricultural practice by considering overall environmental impacts, has gained momentum with the recent improvements in the remote sensing area. The objective of this letter, as a part of precision farming, is to implement Biologische Bundesanstalt, Bundessortenamt und CHemische Industrie (BBCH) scale assignment in plant growth monitoring by means of SAR. The proposed approach copes with structural heterogeneity in agricultural fields by grouping together similar morphologies. For this, densely cultivated paddy rice fields are analyzed using TerraSAR-X (TSX) co-polar SAR data. For generating structurally similar groups, K-means clustering is used in a polarimetric feature vector space, which is composed of backscattering intensities and polarimetric phase differences. This step is followed by a preliminary classification approach based on the temporal separability of the explanatory parameters. In the last step of the proposed methodology, assigned classes are updated based on the biological principles that are followed in rice cultivation. This letter provides the results of the proposed algorithm and compares them to the standard threshold-based approach in two independent agricultural areas. The results show the superiority of the feature-clustering-based classification compared with the standard approach in handling field heterogeneity.

Geostatistical approach for management of soil nutrients with special emphasis on different forms of potassium considering their spatial variation in intensive cropping system of West Bengal, India.

A large part of precision agriculture research in the developing countries is devoted towards precision nutrient management aspects. This has led to better economics and efficiency of nutrient use with off-farm advantages of environmental security. The keystone of precision nutrient management is analysis and interpretation of spatial variability of soils by establishing management zones. In this study, spatial variability of major soil nutrient contents was evaluated in the Ghoragacha village of North 24 Parganas district of West Bengal, India. Surface soil samples from 100 locations, covering different cropping systems of the village, was collected from 0 to 15 cm depth using 100×100 m grid system and analyzed in the laboratory to determine organic carbon (OC), available nitrogen (N), phosphorus (P), and potassium (K) contents of the soil as well as its water-soluble K (KWS), exchangeable K (KEX), and non-exchangeable forms of K (KNEX). Geostatistical analyses were performed to determine the spatial variation structure of each nutrient content within the village, followed by the generation of surface maps through kriging. Four commonly used semivariogram models, i.e., spherical, exponential, Gaussian, and linear models were fitted to each soil property, and the best one was used to prepare surface maps through krigging. Spherical model was found the best for available N and P contents, while linear and exponential model was the best for OC and available K, and for KWS and KNEK, Gausian model was the best. Surface maps of nutrient contents showed that N content (129-195 kg ha(-1)) was the most limiting factor throughout the village, while P status was generally very high ( 10-678 kg ha(-1)) in the soils of the present village. Among the different soil K fractions, KWS registered the maximum variability (CV 75%), while the remaining soil K fractions showed moderate to high variation. Interestingly, KNEX content also showed high variability, which essentially indicates reserve native K exploitation under intensive cultivation. These maps highlight the necessity of estimating the other soil K fractions as well for better understanding of soil K supplying capacity and K fertilization strategy rather than the current recommendations, based on the plant-available K alone. In conclusion, the present study revealed that the variability of nutrient distribution was a consequence of complex interactions between the cropping system, nutrient application rates, and the native soil characteristics, and such interactions could be utilized to develop the nutrient management strategies for intensive small-holder system.

Delineation of management zones using an active canopy sensor for a tobacco field

Recent precision agriculture research has focused on the utilization of management zones (MZs) as a method for variable-rate fertilizer applications. This study tries to make use of an active canopy sensor to define MZs in contrast with the ones using soil properties. To achieve this objective, a 20-m grid-sampling scheme was imposed on the field with 101 points, and we collected soil samples from the top 20cm at each point before tobacco planting in April 2013. The normalized difference vegetative index (NDVI) data at five growth stages of the tobacco growth cycle were measured by using a GreenSeeker handheld crop sensor at the location of each sample point. Yield mapping was carried out at harvest. Through stepwise multiple analysis, we identified the key yield-limiting factors of soil properties and stages of NDVI measured respectively. For soil property data, OM, AP and Fe were used for yield-based MZs while for canopy parameters, NDVI during knee-high and flowering were utilized. Fuzzy c-means clustering algorithm was performed in order to determine the optimal number of the clusters. Results showed that the optimal number of MZs with regard to the two methods were both five and variance analysis indicated the heterogeneity of soil properties and yield among them. Consequently, it is feasible to use an active canopy sensor to delineate management zones for tobacco-planting fields.