Scott A. Shearer

CLASSIFICATION OF WEED SPECIES USING COLOR TEXTURE FEATURES AND DISCRIMINANT ANALYSIS

The environmental impact of herbicide utilization has stimulated research into new methods of weed control, such as selective herbicide application on highly infested crop areas. This research utilized the Color Co-occurrence Method (CCM) to determine whether traditional statistical discriminant analysis can be used to discriminate between six different classes of groundcover. The weed species evaluated were giant foxtail, crabgrass, common lambsquarter, velvetleaf, and ivyleaf morningglory, along with a soil image data set. The between species discriminant analysis showed that the CCM texture statistics procedure was able to classify between five weed species and soil with an accuracy of 93% using hue and saturation statistics, only. A significant accomplishment of this work was the elimination of the intensity texture features from the model, which reduces computational requirements by one-third.

Backpropagation neural network design and evaluation for classifying weed species using color image texture

Color co-occurrence method (CCM) texture statistics were used as input variables for a backpropagation (BP) neural network weed classification model. Thirty-three unique CCM texture statistic inputs were generated for 40 images per class, within a six class data set. The following six classes were studied: giant foxtail, large crabgrass, common lambsquarter, velvetleaf, ivyleaf morningglory, and clear soil surface. The texture data was used to build six different input variable models for the BP network, consisting of various combinations of hue, saturation, and intensity (HSI) color texture statistics. The study evaluated classification accuracy as a function of network topology, and training parameter selection. In addition, training cycle requirements and training repeatability were studied. The BP topology evaluation consisted of a series of tests on symmetrical two hidden-layer network, a test of constant complexity topologies, and tapered topology networks. The best symmetrical BP network achieved a 94.7% classification accuracy for a model consisting of 11 inputs, five nodes at each of the two hidden layers and six output nodes (11 ×5 ×5 ×6 BP network). A tapered topology ( 11 ×12 ×6 ×6 BP network) out performed all other BP topologies with an overall accuracy of 96.7% and individual class accuracies of 90.0% or higher.

PERFORMANCE ASSESSMENT AND MODEL DEVELOPMENT OF A VARIABLE–RATE, SPINNER–DISC FERTILIZER APPLICATOR

The popularity of spinner spreaders for application of granular fertilizers and agricultural lime, along with increased interests in variable–rate technology (VRT), has raised concern about application accuracy and distribution of these spreaders. This investigation was undertaken to assess the application distribution of a VRT spinner fertilizer spreader. Application distribution was assessed using a matrix of collection pans and following test procedures outlined in ASAE Standard S341.2. Uniform and variable–rate tests were performed to characterize the application variability of the spreader and to test the effect of rate changes via GPS control. Uniform and variable–rate application models were developed from the collected data. A sigmoidal function was used to describe increasing application rate changes, while a linear function described decreasing rate changes. Average transverse distribution patterns were used to model both high and low application rates. The resulting models were then compared to the actual distributions. The model was found to do a good job of characterizing uniform and variable–rate application patterns and therefore may be suitable for simulating variable–rate application errors.

Effect of anatomical fractionation on the enzymatic hydrolysis of acid and alkaline pretreated corn stover

Due to concerns with biomass collection systems and soil sustainability there are opportunities to investigate the optimal plant fractions to collect for conversion. An ideal feedstock would require a low severity pretreatment to release a maximum amount of sugar during enzymatic hydrolysis. Corn stover fractions were separated manually and analyzed for glucan, xylan, acid soluble lignin, acid insoluble lignin, and ash composition. The stover fractions were also pretreated with either 0%, 0.4%, or 0.8% NaOH for 2 h at room temperature, washed, autoclaved and saccharified. In addition, dilute sulfuric acid pretreated samples underwent simultaneous saccharification and fermentation (SSF) to ethanol. In general, the two pretreatments produced similar trends with cobs, husks, and leaves responding best to the pretreatments, the tops of stalks responding slightly less, and the bottom of the stalks responding the least. For example, corn husks pretreated with 0.8% NaOH released over 90% (standard error of 3.8%) of the available glucan, while only 45% (standard error of 1.1%) of the glucan was produced from identically treated stalk bottoms. Estimates of the theoretical ethanol yield using acid pretreatment followed by SSF were 65% (standard error of 15.9%) for husks and 29% (standard error of 1.8%) for stalk bottoms. This suggests that integration of biomass collection systems to remove sustainable feedstocks could be integrated with the processes within a biorefinery to minimize overall ethanol production costs.

Reducing Pesticide Over-Application with Map-Based Automatic Boom Section Control on Agricultural Sprayers

The use of precision agriculture technologies such as automatic guidance and map-based automatic boom section control on agricultural sprayers has increased substantially over the past few years. The purpose of these systems is to decrease pesticide application inaccuracy by reducing pass-to-pass overlaps or by turning boom sections off when the boom passes over previously covered areas or beyond field boundaries. The objectives of this study were to compare areas treated by a sprayer before and after the addition of an automatic boom section control system and determine if there was a relationship between field shape factors, specifically perimeter-to-area (P/A) ratio, and pesticide over-application. Coverage files were downloaded from a self-propelled agricultural sprayer with a 24.8 m boom divided into five manually controlled sections. Prior to the subsequent cropping season, automatic boom section control (seven sections) was installed on the sprayer. Application inaccuracy was calculated for the 21 study fields by comparing the actual sprayed area (based on boom section control state: on/off) versus the total projected field area. A comparison between the average percent over-application from fields during season one (12.4%) and season two (6.2%) revealed a significant reduction in off-target application. This reduction represents savings to agricultural producers, potentially justifying the purchase and implementation of this technology. Further analysis indicated an increasing trend in over-application for manual and automatic boom section control as the P/A ratio increased for the study fields. Over-application increased at a greater rate with manual boom section control, which suggests that as field inclusions (grassed waterways or other obstacles) increase, automatic boom section control will provide a greater opportunity for producers to reduce these errors.

Distribution pattern variability of granular VRT applicators

Granular applicators equipped with variable-rate technology (VRT) have gained popularity in recent years as a result of increased interest in variable-rate application. The purpose of this investigation was to characterize distribution patterns at varying rates for different granular applicators. Uniform-rate (UR) tests were conducted to assess the accuracy of variable-rate application from four granular applicators: two spinner-disc spreaders (A and B), and two pneumatic applicators (C and D). Pattern results indicated a consistent triangular pattern for spinner spreader B and consistent patterns for the pneumatic applicators (C and D). However, applicator D produced pattern variations at the center and right side. Simulated overlap analysis generated CVs <20% for applicators B and C. Applicator A performed well at the two lower rates (CVs <19%) but not at the highest rate (CV = 27%). Pattern unevenness for applicator D produced CVs between 25% and 34%. The spinner-disc spreaders over-applied, while the pneumatic applicators under-applied at the margins, suggesting an adjustment to the effective swath spacing or spinner-disc speed is needed to improve application accuracy. Further, overlap plots indicated pattern variability even when acceptable CVs were attained for applicators B and C. Therefore, it is recommended that CVs accompany simulated overlap pattern plots to ensure proper calibration of VRT equipment. Swath spacing analysis indicated that three of the four applicator spacings could be changed from the recommended value to improve application uniformity. Pattern comparisons showed that pattern shifts occurred for applicator A (P = 0.0092) with increasing application rate but not for applicators B, C, and D. These results demonstrate potential application errors with VRT and the need for proper calibration to maintain acceptable performance. Further, this investigation demonstrates the need for a VRT equipment testing standard.

An overview of current and potential applications of thermal remote sensing in precision agriculture

Abstract Precision agriculture (PA) utilizes tools and technologies to identify in-field soil and crop variability for improving farming practices and optimizing agronomic inputs. Traditionally, optical remote sensing (RS) that utilizes visible light and infrared regions of the electromagnetic spectrum has been used as an integral part of PA for crop and soil monitoring. Optical RS, however, is slow in differentiating stress levels in crops until visual symptoms become noticeable. Surface temperature is considered to be a rapid response variable that can indicate crop stresses prior to their visual symptoms. By measuring estimates of surface temperature, thermal RS has been found to be a promising tool for PA. Compared to optical RS, applications of thermal RS for PA have been limited. Until recently (i.e., before the advancement of low cost RS platforms such as unmanned aerial systems (UAVs)), the availability of high resolution thermal images was limited due to high acquisition costs. Given recent developments in UAVs, thermal images with high spatial and temporal resolutions have become available at a low cost, which has increased opportunities to understand in-field variability of crop and soil conditions useful for various agronomic decision-making. Before thermal RS is adopted as a routine tool for crop and environmental monitoring, there is a need to understand its current and potential applications as well as issues and concerns. This review focuses on current and potential applications of thermal RS in PA as well as some concerns relating to its application. The application areas of thermal RS in agriculture discussed here include irrigation scheduling, drought monitoring, crop disease detection, and mapping of soil properties, residues and tillage, field tiles, and crop maturity and yield. Some of the issues related to its application include spatial and temporal resolution, atmospheric conditions, and crop growth stages.

Whole farm analysis of automatic section control for agricultural machinery

Automatic section control was analyzed in a whole farm decision-making framework when implemented on an agricultural sprayer and/or planter. In addition, various field types and navigational scenarios were examined to determine their impact on profitability. It was determined that automatic section control increased net returns under all scenarios; up to

CONTROLLER AREA NETWORK BASED DISTRIBUTED CONTROL FOR AUTONOMOUS VEHICLES

36/ha. This investigation highlighted the importance of considering field size in addition to field shape as well as initial navigational scenarios when determining the profitability of automatic section control.

Grain Yield Monitor Flow Sensor Accuracy for Simulated Varying Field Slopes

The goal of this project was to evaluate the potential of a controller area network (CAN bus) to be used as the communication network for a distributed control system on an autonomous agricultural vehicle. The prototype system utilized microcontroller-driven nodes to act as control points along a CAN bus. Messages were transferred to the steering, transmission, and hitch control nodes via a task computer. The task computer utilized global positioning system data to generate appropriate control commands. Laboratory and field testing demonstrated that each of the control nodes could function simultaneously over the CAN bus. Results showed that the task computer adequately applied a feedback control model to the system and achieved guidance accuracy levels well within the desired range. Testing also demonstrated the systems ability to complete normal field operations, such as headland turning and implement control. ver the past several years, technology has contin- ued to play an increasing role in agriculture. The industry has recently seen the advent and develop- ment of many types of automated vehicles ranging from planters to sprayers to harvesters. These vehicles have all sustained different levels of automation. Some were capa- ble of fully autonomous field operations, while others were developed for specific control operations such as autosteer- ing (Reid et al., 2000). Commercialization of these technolo- gies has come at a substantial cost to the farmer. Autosteer systems range in cost from