Ralph B. Brown

Site-specific weed management: sensing requirements— what do we need to see?

Abstract Automated detection and identification of weeds in crop fields is the greatest obstacle to development of practical site-specific weed management systems. Research progress is summarized for two different approaches to the problem, remote sensing weed mapping and ground-based detection using digital cameras or nonimaging sensors. The general spectral and spatial limitations reported for each type of weed identification system are reviewed. Airborne remote sensing has been successful for detection of distinct weed patches when the patches are dense and uniform and have unique spectral characteristics. Identification of weeds is hampered by spectral mixing in the relatively large pixels (typically larger than 1 by 1 m) and will not be possible from imagery where weed seedlings are sparsely distributed among crop plants. The use of multispectral imaging sensors such as color digital cameras on a ground-based mobile platform shows more promise for weed identification in field crops. Spectral features plus spatial features such as leaf shape and texture and plant organization may be extracted from these images. However, there is a need for research in areas such as artificial lighting, spectral band requirements, image processing, multiple spatial resolution systems, and multiperspective images.

Fuzzy control system for peanut roasting

Abstract A fuzzy control system was developed for continuous, cross-flow peanut roasting. A roasting kinetic model based on the combined dynamics of heat transfer and browning reactions was developed. A combination feedforward–feedback scheme was implemented in application software developed by this research group for fuzzy rule-based control. Inputs to the control system included numeric values from process sensors as well as linguistic observations by operators. The control system was tested on a pilot-scale roaster and it successfully maintained roasted peanut colour within an acceptable range despite disturbances in roaster bed depth, roasting air temperature and colour setpoint.

Evaluation of site-specific weed management using a direct-injection sprayer

Abstract Targeting weed patches for site-specific herbicide applications potentially represents cost savings for operators, reduction in environmental herbicide effects, and increased efficiency of weed control. An experiment was initiated in a no-till corn field in Ontario, Canada, in 1998 and was continued in rotation with no-till soybeans in 1999. Weeds were intensively scouted, and distribution maps of the most common weeds (field horsetail, spiny sowthistle, dandelion, and common lambsquarters) were generated for both years. A prescription map for each plot was made using the weed density maps. Treatment decisions were based on a weed threshold value of 1 shoot m−2. Four herbicide treatments were compared: a conventional broadcast, a site-specific application targeting weed patches only, and two combinations of broadcast and site-specific applications. Treatments were applied using a direct-injection sprayer. Efficacy of weed control and yield were compared among treatments. In 1998 and 1999 there were no differences in the level of weed control or yield among treatments. The average percent area sprayed was reduced as much as 26% in the site-specific treatment in 1998 and up to 59% in the site-specific and broadcast combination treatments in 1999. For those species present in the field, patches ranged from highly aggregated to completely random, and patch stability ranged from very stable to very unstable over the 2 yr. Nomenclature:Common lambsquarters, Chenopodium album L. CHEAL; corn, Zea mays L.; dandelion, Taraxacum officinale Weber. TAROF; field horsetail, Equisetum arvense (L.) EQUAR; soybean, Glycine max (L.) Merr.; spiny sowthistle, Sonchus asper (L.) Hill. SONAS.

Impact of common ragweed (Ambrosia artemisiifolia) aggregation on economic thresholds in soybean

Abstract One approach to site-specific weed control is to map weeds within a field and then divide the field area into smaller grid units. The decision to apply a herbicide to individual grid units, or decision units, is made by using yield loss models to establish an economic threshold level. However, decision units often contain weed populations with aggregated distributions. Many yield loss models have not considered this because experiments dealing with weed–crop competition typically assume uniform weed distributions. Therefore, these models may overestimate yield losses. Field experiments conducted in 1999 and 2000 compared the effects of common ragweed having a uniform distribution vs. an aggregated distribution on soybean seed yield, moisture content, and dockage. Field experiment data were used to calculate and compare economic thresholds for both distributions. Economic thresholds that considered drying costs and dockage also were compared. There was no significant difference in I parameters (yield loss as density approaches zero) between the two ragweed distributions in either year. Seed moisture content and dockage increased with increasing common ragweed densities, but increases were not significant at the break-even yield loss level. Economic threshold values were similar for both distributions with differences between aggregated and uniform of 0.14 and 0.01 plants m−2 in 1999 and 2000, respectively. The economic threshold values were reduced by 0.01 to 0.06 plants m−2 when drying costs and dockage were considered. Nomenclature: Common ragweed, Ambrosia artemisiifolia L. AMBEL; soybean, Glycine max L. ‘First Line 2801R’.

Remote Sensing of Soil Moisture in Vineyards Using Airborne and Ground-Based Thermal Inertia Data

Thermal remote sensing of soil moisture in vineyards is a challenge. The grass-covered soil, in addition to a standing grape canopy, create complex patterns of heating and cooling and increase the surface temperature variability between vine rows. In this study, we evaluate the strength of relationships between soil moisture, mechanical resistance and thermal inertia calculated from the drop of surface temperature during a clear sky night over a vineyard in the Niagara region. We utilized data from two sensors, an airborne thermal camera (height ≈ 500 m a.g.l.) and a handheld thermal gun (height ≈ 1 m a.g.l.), to explore the effects of different field of views and the high inter-row temperature variability. Spatial patterns of soil moisture correlated more with estimated thermal inertia than with surface temperature recorded at sunrise or sunset. Despite the coarse resolution of airborne thermal inertia images, it performed better than estimates from the handheld thermal gun. Between-row variation was further analyzed using a linear mixed-effects model. Despite the limited spatial variability of soil properties within a single vineyard, the magnitudes of the model coefficients for soil moisture and mechanical resistance are encouraging indicators of the utility of thermal inertia in vineyard management.

Fuzzy predictor for fermentation time in a commercial brewery

Abstract A study was conducted at a commercial brewery to investigate the feasibility of an expert-system control strategy that used fuzzy logic to make inferences based on uncertain process information. The predictor made an initial estimate of fermentation time based on the yeast variables (pitching rate, pitched volume and viability). The predicted fermentation times were within 24 h of the actual fermentation time for 9 batches in the validation data set of 13 process histories. A second rule set followed the pH and specific gravity during fermentation and predicted the level of vicinal diketone (VDK). When the VDK level reached a threshold level, a third rule set updated the estimate of time required to complete the fermentation. Seven data sets were used to validate the predictions based on VDK level and 4 estimates were within 24 h of the recorded fermentation release times. All 7 predictions were within 32 h of the actual release time and 6 of the 7 estimates indicated that the batches were held longer than necessary.

Visible-Near Infrared Reflectance Spectroscopy for Nondestructive Analysis of Red Wine Grapes

Composite samples of intact grape berries were collected weekly from veraison until harvest. Each sample comprised ~400 berries selected following the preharvest row sampling protocol specified by the vineyard manager. The grape cultivars and corresponding number of samples of each collected in 2009 and 2010 were as follows: Cabernet Sauvignon (43, 36), Cabernet franc (83, 80), and Syrah (38, 36). Reflectance spectra for the composite samples in a wavelength range of 350 to 850 nm were collected with a diode array spectrometer. Chemical analyses for soluble solids content, Brix, pH, titratable acidity (TA), total phenols, and total anthocyanins were carried out for all samples. Chemometric calibrations for corresponding reflectance data were developed using trained partial least squares regression models with several preprocessing methods (smoothing, normalization, differentiation) and subjected to variable selection by recursive feature elimination. Trained models were externally validated with data from the alternate year. Best performing models for Brix, pH, TA, phenols, and anthocyanins in 2009 had root mean square errors (RMSEP) of 0.65, 0.05, 0.59 g/L, 31.2 mg/L, and 75 mg/L, respectively, with corresponding R2 values of 0.84, 0.58, 0.56, 0.27 and 0.65. The best 2010 models had RMSEP of 0.65, 0.05, 0.86 g/L, 27.9 mg/L, and 111 mg/L, respectively, with corresponding R2 values of 0.89, 0.81, 0.58, 0.25, and 0.17. The 2009 calibrations were used for estimating Brix and pH from spectral data of the samples collected in the next growing season and yielded RMSEP performance of 0.87 and 0.05 and R2 values of 0.71 and 0.56, respectively. Principal component analysis decomposition of 2009 and 2010 reflectance data showed similarities in the resultant loadings, indicating a similar underlying data structure.

Separating near surface thermal inertia signals from a thermal time series by standardized principal component analysis

Principal component analysis has been applied to remote sensing data to identify spatiotemporal patterns in a time series of images. Thermal inertia is a surface property that relates well to shallow surface thermal and physical properties. Mapping thermal inertia requires quantifying surface energy balance components and soil heat flux, both of which are difficult to measure remotely. This article describes a method to map soil thermal inertia using principal component analysis applied to a time series of thermal infrared images and it also assesses how sensitive this method is to the time intervals between images. Standardized principal component analysis (SPCA) was applied to thermal infrared images captured at half-hour intervals during a complete diurnal cycle. Shallow surface thermal properties accounted for 45%, 82% and 66% of the spatiotemporal variation in surface temperature observed during the heating phase, cooling phase and over the total diurnal cycle respectively. The remaining 55%, 18% and 34% of the variation was attributed to transient effects such as shadows, surface roughness and background noise. Signals related to thermal inertia explained 18% of total variation observed in a complete diurnal cycle and 7% of variation in the cooling series. The SPCA method was found useful to separate critical information such as timing and amplitude of maximum surface temperature variation from delays related to differential heating induced by micro-topography. For the field conditions experienced in this study, decreased temporal resolution when sampling intervals were greater than an hour significantly reduced the quality of results.

Using remote sensing to understand Pinot noir vineyard variability in Ontario

The study objective was to determine whether multispectral high spatial resolution airborne imagery could be used to segregate zones in Pinot noir vineyards to target highest quality fruit for premium winemaking. We hypothesized that remotely sensed data would be correlated with vine size and leaf water potential (Ψ), and by extension with yield and berry composition. In 2008-2009, multispectral (blue, green, red, near-infrared) airborne images were acquired from four Ontario Pinot noir vineyards (four dates, 2008; three, 2009), with the final flight date near veraison. A process was developed to extract information from sentinel vine locations, and to calculate normalized difference vegetation index (NDVI). Data were extracted at 1 × 1, 3 × 3, and 5 × 5 pixel re-sampling rates to test for ideal image resolution. A method was developed to mask non-vine pixels to simplify qualitative assessment of images. The 3 × 3 pixel re-sampling provided most useful information. In 2008, 3 × 3 re-sampling NDVI correlated with (r-values; p < 0.0001): berry pH (-0.48), soluble solids (-0.43), vine size (0.46), anthocyanins (-0.65), colour (-0.58), and soil clay and sand content (-0.55, 0.55). In 2009, mean 3 × 3 re-sampling NDVI correlated with (r-values; p < 0.0001): anthocyanins (0.49), soil moisture (-0.89), and soil clay and silt content (-0.75, 0.83). No clear trends in correlations existed between vegetation indices vs. vine size, anthocyanins, phenolics, or soil moisture throughout the growing season in either vintage. Masked images proved effective for viewing spatial trends in airborne images without full data extraction. Qualitative similarities existed between maps of vineyard and grape composition variables vs. maps of extracted data and masked images. Remote sensing may be useful to determine colour or phenolic potential of grapes, in addition to vine water status, yield, and vine size. This study was unique by employing remote sensing in cover-cropped vineyards and thereafter using protocols for excluding spectral reflectance contributed by inter-row vegetation.

Forced-Air Drying of Ginseng Root: Pilot-Scale Control System for Three-Stage Process

Pilot-scale experiments were conducted to evaluate a three-stage temperature control system (38–50–38°C) for ginseng root drying. Feedback for determining temperature change points was provided by measuring root shrinkage in digital images of the root bed. Shrinkage kinetics observed during the first stage were used to predict the drying rate for the third stage and an exponential model was used to predict the time required for the roots to reach an average moisture content of 0.11 g/g (db). Based on experimental tests, the root mean square error for the final moisture was 0.025 g/g (db).