Search

A Gabor wavelets based technique was investigated as a potential tool for scene classification (into one of bare patch, plant, or weed) for its ultimate utility in site-specific application of agrochemicals in wild blueberry fields. Images were gathered from five sites located in central Nova Scotia, Canada. Gabor wavelet features extracted from these images were used to classify scenes according to visually determined classes using step-wise linear discriminant analysis. For individual fields, classification accuracy attained ranged between 87.9% and 98.3%; selected Gabor features ranged between 27 and 72; contextual accuracy for herbicide ranged between 67.5% and 96.7%, and contextual accuracy for fertilizer ranged between 63.6% and 97.1%. The pooled scenes yielded a classification accuracy of 81.4%, and contextual accuracy figures of 61.1% and 73.1% for herbicide and fertilizer, respectively, with selected Gabor features of 36. Calibrations based on LDA coefficients from the pooled scenes could help avoid the need to re-calibrate for each field, whereas those based on individual field LDA coefficients could improve accuracy, hence enable saving on expensive agrochemicals.

The study extracts representative features to train a model with supervised machine learning (ML) to detect powdery mildew (Sphaerotheca macularis f. sp. fragariae) on the strawberry leaves. Powdery mildew (PM) is a fungal disease that greatly affects the production of strawberry and usually infects under conditions of warming temperatures and high humidity. In this research, we report robust models to detect PM using image processing and ML technologies. Three feature extraction techniques (histogram of oriented gradients; HOG, speeded-up robust features; SURF, and gray level co-occurrence matrix; GLCM) and two supervised ML (artificial neural network; ANN and support vector machine; SVM) were implemented using MATLAB. Images were augmented to 1016 images using a four different angle rotation technique to simulate strawberry leaf bundles in the real field. The classification accuracy (CA) to detect PM was highest at 94.34% with a combination of ANN and SURF with 908 × 908 image resolution and with SVM and GLCM at 88.98% with 908 × 908 image resolution. In terms of the extraction time for real-time processing, HOG takes the shortest time to extract features in both ANN and SVM.

With being rapid increasing population in worldwide, the need for satisfactory level of crop production with decreased amount of agricultural lands. Machine vision would ensure the increase of crop production by using an automated, non-destructive and cost-effective technique. In last few years, remarkable results have been achieved in different sectors of agriculture. These achievements are integrated with machine learning techniques on machine vision approach that cope with colour, shape, texture and spectral analysis from the image of objects. Despite having many applications of different machine learning techniques, this review only described the statistical machine learning technologies with machine vision systems in agriculture due to broad area of machine learning applications. Two types of statistical machine learning techniques such as supervised and unsupervised learning have been utilized for agriculture. This paper comprehensively surveyed current application of statistical machine learning techniques in machine vision systems, analyses each technique potential for specific application and represents an overview of instructive examples in different agricultural areas. Suggestions of specific statistical machine learning technique for specific purpose and limitations of each technique are also given. Future trends of statistical machine learning technology applications are discussed.

Strawberry cropping system relies heavily on proper disease management to maintain high crop yield. Powdery mildew, caused by Sphaerotheca macularis (Wall. Ex Fries) is one of the major leaf diseases in strawberry which can cause significant yield losses up to 70%. Field scouts manually walk beside strawberry fields and visually observe the plants to monitor for powdery mildew disease infection each week during summer months which is a laborious and time-consuming endeavor. The objective of this research was to increase the efficiency of field scouting by automatically detecting powdery mildew disease in strawberry fields by using a real-time machine vision system. A global positioning system, two cameras, a custom image processing program, and a ruggedized laptop computer were utilized for development of the disease detection system. The custom image processing program was developed using color co-occurrence matrix-based texture analysis along with artificial neural network technique to process and classify continuously acquired image data simultaneously. Three commercial strawberry field sites in central Nova Scotia were used to evaluate the performance of the developed system. A total of 36 strawberry rows (~1.06 ha) were tested within three fields and powdery mildew detected points were measured manually followed by automatic detection system. The manually detected points were compared with automatically detected points to ensure the accuracy of the developed system. Results of regression and scatter plots revealed that the system was able to detect disease having mean absolute error values of 4.00, 3.42, and 2.83 per row and root mean square error values of 4.12, 3.71, and 3.00 per row in field site-I, field site-II, and field site-III, respectively. The slight deviation in performance was likely caused by high wind speeds (&gt, 8 km h&minus, 1), leaf overlapping, leaf angle, and presence of spider mite disease during field testing.

Wild blueberry crop yields are dependent on heavy agrochemical applications to control weeds competing with the crop. Goldenrod is a creeping herbaceous perennial weed that occurred in 90% of wild blueberry fields surveyed in Nova Scotia. A colour co-occurrence matrices (CCMs) based algorithm can be used for spot-application of herbicide on goldenrod within wild blueberry fields. The objective of this study was to analyse the effect of different parameters on computational complexity and goldenrod identification accuracy of CCMs in wild blueberry cropping system. An image acquisition graphical user interface (GUI) and CCMs based textural analysis algorithm were developed in Microsoft Visual® C# programming language. The GUI was used to acquire 2244 area of interest images along with a set of 2244 full frame images from two different wild blueberry fields in central Nova Scotia, Canada. Six image intensity levels, seven image sizes, and three CCMs were used to for the study. The results indicated that intensity levels and image size significantly influenced the computational requirements of CCMs. Images with 256 or 128 intensity levels could be used as these levels correctly classified 94% and 89% of test observations respectively. The processing times were increased from 535 μs to 10,650 μs and 9864 μs to 63,750 μs as the intensity increased from 8 to 256 levels and image size increased from 16 × 16 to 1024 × 1024 pixels, respectively. The time required for textural feature extraction was not statistically significant for different image sizes used in this study. Overall, the results indicated that intensity levels of 128 or 256, a unit image size of 128 × 128 pixels, and saturation colour plane alone or in combination with hue can help to minimise the processing burden without compromising the classification accuracy for real-time applications.

An essential part of the wild blueberry cropping system is the proper management of agrochemical inputs including herbicides, fungicides and insecticides. A machine vision system was developed and mounted on the rear sprayer boom 0.18 m in front of the sprayer nozzles capable of targeting the agrochemical application on an as-needed basis. The three-point hitch mounted sprayer featured 27 nozzles over a 13.7 m boom width and a storage tank capacity of 1135 l. Nine digital color cameras continually take images in real-time while computer software processes the images in 0.15 s to determine the target locations where the nozzles open and spray at speeds up to 1.77 m s−1. Two wild blueberry fields in central Nova Scotia were selected for smart sprayer performance testing with spot-application (SA) of agrochemical as compared to control and uniform application techniques. Chateau® herbicide was applied in a field with an infestation of hair cap moss. Spray droplet comparison showed moss patches were properly targeted using the smart sprayer. SA provided the same coverage performance as compared to uniform on the moss targets with herbicide application savings of 78.5% using the smart sprayer. Harvestable yield results were similar for all application tracks. TruPhos Magnesium and ZincMax foliar fertilizers were tank mixed with Bravo® and Proline® fungicides and applied to compare the difference of SA, control and uniform application. Results showed SA of foliar fertilizer and fungicide led to less premature leaf drop and increased the blueberry stem height, number of branches, stem diameter and fruit buds. SA of foliar fertilizer and fungicide also increased the percent of healthy wild blueberry plants by 57.8% and the harvestable yield by 137.8%. Fungicide application savings using the smart sprayer for SA were 11.6%.

Northeastern North America is the world’s leading producer of wild blueberry. Ripening of wild blueberry is the leading factor for fruit quality. Currently, there are no protocols available for the farming community related to wild blueberry fruit ripening and maturity. A nondestructive, rapid, and reliable digital photography technique could be used to examine the ripening of wild blueberries for appropriate harvesting time. Two wild blueberry fields were selected to examine the berry ripening levels using digital photographic techniques at different time of harvest (early, middle, and late seasons). The fields were divided into four blocks and each block was further divided into three classes of times of harvest. Fruit images from each block were acquired and processed to count blue pixels from each image. A significant correlation was found between percentage of blue pixels and actual fruit yield in Field A (R2 = 0.96; P < 0.001) and Field B (R2 = 0.97; P < 0.001). The results also indicated that the effect of time of harvest on fruit yield was significant and fruit yield increased gradually during early harvesting, reached maximum during mid-season, and then started to decrease in late harvesting. Results indicated that 90% of green-berries had turned blue at the end of middle season compared to early season (58%). Based on the results of this study, optical analysis could help to keep fruit quality by optimizing appropriate harvesting time of wild blueberries. It is also suggested that the optimum time to harvest wild blueberries is middle season to ensure high fruit yield and quality. Keywords: Blue pixels, Fruit yield, Harvesting season, Wild blueberry.

Spatial variations in wild blueberry plant height pose a serious challenge for the operator to maintain the optimum harvester picking head height. Harvester head adjustment based on plant height increases harvestable fruit yield and quality while preventing plant pulling. An advanced Ultrasonic Plant Height Measurement System II (UPHMS II) was developed and incorporated into a commercial mechanical harvester. The developed system consisted of three ultrasonic sensors covering the width of the harvester head, a real-time kinematics global positioning system (RTK-GPS), custom built software, and a rugged computer. The custom software acquired and processed the ultrasonic sensing data in real-time during mechanical harvesting. Four wild blueberry fields were selected in central Nova Scotia to evaluate the performance of the developed system. Forty eight experimental plots were randomly constructed within four fields and wild blueberry plant heights were recorded manually prior to harvest. The UPHMS II was tested and evaluated to estimate plant height. The manual plant height measurements were compared with ultrasonically sensed data to ensure the accuracy of the developed system. A previous system (UPHMS I) comprising of one ultrasonic sensor was also tested and evaluated. Results of regression and scatter plots revealed that the UPHMS II was able to sense plant height in real-time more accurately with root mean square error (RMSE) of 1.7 cm when compared with the UPHMS I (RMSE = 5.7 cm). The UPHMS II equipped with three sensors covered the complete width of the harvester (0.91 m), which showed higher accuracy compared to the UPHMS I.

High wind constraints during day time agrochemical spraying has pushed the wild blueberry producers to apply agrochemicals during the early morning, evening or after dark, to avoid drift problems due to low wind conditions. The objective of this study was to develop an artificial light source system combined with a smart sprayer comprising of a digital camera-based sensing system to allow cameras to detect target areas (weed, plant or bare soil) in real-time for accurate application of agrochemicals in low light conditions. After testing and evaluation of different light sources, a rugged light source system equipped with polystyrene diffuser sheets was constructed to provide an even distribution of light across the entire 12.2 m machine vision sensor boom. Distribution of artificial light underneath the sensing boom at zero ambient light was examined by recording the light intensity at 0.15 m spacing on the ground under the camera boom using a lux meter. Results of light distribution revealed that the Magnafire ® 70 W high intensity discharge (HID) lights provided wide angle of even light illumination, high intensity and rugged construction. A wild blueberry field was selected in central Nova Scotia, Canada, and a test track was made to evaluate the performance of the artificial light source system to apply agrochemicals on a spot-specific basis under low natural light conditions. A real-time kinematics-global positioning system (RTK-GPS) was used to map the boundary of the test track, selected bare soil areas, weed areas and wild blueberry plant areas in the field. Water sensitive papers (WSPs) were placed at randomly selected locations, the smart sprayer was operated under low light conditions, and the percent area coverage (PAC) was calculated. The mean PAC from WSP located in bare soil, weeds and blueberry spots in the track was 5.19%, 27.53%, and 1.74%, respectively. PAC of the WSPs placed in bare soil and blueberry patches were 22.34% and 25.79% lower than in weed patches, respectively. Results reported that the custom developed artificial light source system was accurate enough to detect targets in low light conditions. Additionally, spot-spacing only in weed areas resulted in 65% of chemical saving.

Spatial variability in fruit losses in relation to fruit yield, plant height, and ground slope can help to automate the wild blueberry harvester to improve picking performance. Currently, harvester operators adjust harvester’s head height, ground speed, and revolutions per minute (rpm) manually. This is not only laborious but also stressful for operators, as they encounter spatial variability during harvesting. The goal of this work was to identify the automation potential of the harvester to improve harvestable yield and reduce operator’s stress, keeping in view the spatial variability. Two fields were selected and test plots were constructed to examine the performance of the harvester in five zones of plant height, fruit yield, and ground slope. Fruit yield plant height and ground slope were recorded from each plot manually to examine their impact on total fruit loss. Keywords: Automation, Fruit losses, Spatial variability, Wild blueberry, Zonal analysis.Results confirmed significant variability in fruit yield, plant height, and ground slope. Fruit losses were significantly influenced by the spatial variations. Fruit losses increased with an increase in fruit yield and ground slope during mechanical harvesting. The picking performance of the blueberry harvester was significantly lower in short and very tall plants within selected fields. The dependence of fruit losses on fruit yield, plant height, and ground slope emphasize the need for real-time adjustments in machine operating parameters to improve berry recovery. Based on the results, it is concluded that there is a significant advantage of harvester’s automation to increase profit margins for growers with no additional cost. Keywords: Automation, Fruit losses, Spatial variability, Wild blueberry, Zonal analysis.

The goal of the project was to supply growers with knowledge on how incorporation of machine vision technology can affect the wild blueberry crop, disease pressures, and the overall savings of select agrochemical inputs. A machine vision system was developed and mounted on a rear sprayer boom in front of the sprayer nozzles capable of targeting the agrochemical application on an as-needed basis. Results showed that plants that received the proper fungicide application were less prone to premature leaf drop resulting in larger stem diameters and higher bud counts and harvestable fruit yield. Fungicide application savings using the smart sprayer for spot-application was 12% as compared to a uniform application.

The global production of broiler meat was forecasted to be 97.8 MT in 2019. The cost fluctuations create risks in production. In order to have an effective management system, process uncertainty must be taken into account. This approach considers the process as an interval with fuzzy numbers and, for managing the risks, it uses the variable &alpha, a parameter determined by the manager in an interval between 0 and 1. Then two algorithms, namely the multi-objective imperialist competitive algorithm (MOICA) and multi-objective particle swarm optimization (MOPSO), were compared and applied. Since the process of production has many activities and each activity has possible options, the process does not have a unique solution. Therefore, the objective function and its assigned weights in terms of time, cost, and quality can be applied to select the best solution from those obtained. A vast amount of uncertainty can be observed, and effective management necessitates dealing with these uncertainty issues. The MOPSO algorithm showed better performance than the MOICA algorithm in this problem. Based on fuzzy logic and influenced by the uncertainty condition (&alpha, = 0), time, cost, and quality in the MOPSO and the MOICA algorithms were 1793.8 h, $260,571.7, and 46.66%, and 1792.5 h, $260,585.7, and 51.19%, respectively.

Strawberry is an important fruit crop in Canada but powdery mildew (PM) results in about 30&ndash, 70% yield loss. Detection of PM through an image texture-based system is beneficial, as it identifies the symptoms at an earlier stage and reduces labour intensive manual monitoring of crop fields. This paper presents an image texture-based disease detection algorithm using supervised classifiers. Three sites were selected to collect the leaf image data in Great Village, Nova Scotia, Canada. Images were taken under an artificial cloud condition with a Digital Single Lens Reflex (DSLR) camera as red-green-blue (RGB) raw data throughout 2017&ndash, 2018 summer. Three supervised classifiers, including artificial neural networks (ANN), support vector machine (SVM), and k-nearest neighbors (kNN) were evaluated for disease detection. A total of 40 textural features were extracted using a colour co-occurrence matrix (CCM). The collected feature data were normalized, then used for training and internal, external and cross-validations of developed classifiers. Results of this study revealed that the highest overall classification accuracy was 93.81% using the ANN classifier and lowest overall accuracy was 78.80% using the kNN classifier. Results identified the ANN classifier disease detection having a lower Root Mean Square Error (RMSE) = 0.004 and Mean Absolute Error (MAE) = 0.003 values with 99.99% of accuracy during internal validation and 87.41%, 88.95% and 95.04% of accuracies during external validations with three different fields. Overall results demonstrated that an image texture-based ANN classifier was able to classify PM disease more accurately at early stages of disease development.

In this study, Deep Learning (DL) was used to detect powdery mildew (PM), persistent fungal disease in strawberries to reduce the amount of unnecessary fungicide use, and the need for field scouts. This study optimised and evaluated several well-established learners, including AlexNet, SqueezeNet, GoogLeNet, ResNet-50, SqueezeNet-MOD1, and SqueezeNet-MOD2. Data augmentation was carried out from among 1450 healthy and infected leaf images to prevent overfitting and to consider the various shapes and direction of the leaves in the field. A total of eight clockwise rotations (0°; the original data, 45°, 90°, 135°, 180°, 225°, 270°, and 315°) was performed to generate 11,600 data points. Overall, the six DL algorithms that were used in this study showed on average of >92% in classification accuracy (CA). ResNet-50 gave the highest CA of 98.11% in classifying the healthy and infected leaves; however, considering the computation time, AlexNet had the fastest processing time, at 40.73 s, to process 2320 images with a CA of 95.59%.When considering the memory requirements for hardware deployment, SqueezeNet-MOD2 would be recommended for PM detection on strawberry leaves with a CA of 92.61%.

Wild blueberry is one of the most important fruit crops of Canada that produces more than 50% of the world‘s blueberries. Understanding and predicting the relationships between the machine operating parameters, fruit losses, topographic features, and crop characteristics can aid in better berry recovery during mechanical harvesting. This article suggested a modeling approach for prediction of fruit losses during harvesting using artificial neural network (ANN) and multiple regression (MR) techniques. Four wild blueberry sites were selected and completely randomized factorial (3 x 3) experiments were conducted at each site. One hundred sixty-two plots (0.91 x 3 m) were made at each site, in the path of operating harvester. Total fruit yield and losses were collected from each plot within selected sites. The harvester was operated at specific levels of ground speed (1.20, 1.60, and 2.00 km h-1) and head rotational speed (26, 28, and 30 rpm). The slope, plant height, and fruit zone were also recorded from each plot. The collected data were normalized, and 70% of the data were utilized for calibration, and 30% for validation of developed models using ANN and MR techniques. Results of root mean square error (RMSE) suggested that the tanh-sigmoid transfer function between the hidden layer and output layer was the best fit for this study. The developed models were validated internally and externally and the best performing configurations were identified based on RMSE, coefficient of efficiency, percent variation, and coefficient of determination. Results of scatter plot among the RMSE and epoch suggested that an epoch size (iterative steps) of 15,000 was appropriate to predict fruit losses using ANN approach. Results revealed that the prediction accuracy of MR model was lower (R2 = 0.46; RMSE = 0.14%) than the ANN model (R2 = 0.84; RMSE = 0.075%) for calibration dataset. Results reported that the ANN model predicted fruit losses with higher (R2 = 0.63; RMSE = 0.11%) accuracy when compared with MR model (R2 = 0.37; RMSE = 0.15%) for external validation dataset. Overall, results of this study suggested that the ANN model was able to accurately and reliably predict fruit losses during harvesting. These results can help to identify the factors responsible for fruit losses and to suggest optimal harvesting scenarios to improve harvesting efficiency.

Understanding the relationships between soil properties, topographic features, and their impact on fruit yield is an important component of site-specific nutrient management. This article presents research that investigates the relationship between soil properties and fruit yield to provide a better understanding of their variation with respect to change in slope within selected wild blueberry fields. Two wild blueberry fields were selected and completely randomized experiments were constructed at each site. The selected sites were divided into five slope categories: 0 16° (Category 5). Soil samples were collected from each slope category and analyzed for pH, texture, electrical conductivity (EC), soil organic matter (SOM), and inorganic nitrogen (NH4+-N, NO3--N). The slope, volumetric moisture content (?v), and fruit yield data were also collected from each slope category. Soil variables, slope and fruit yield data were replicated ten times in each slope category.

An automated sensing and control system (hardware and software) was developed for real-time spot-application of granular fertilizer in mowed wild blueberry fields. The custom hardware system was incorporated into a commercial pneumatic granular fertilizer spreader. Custom software for the sensing and control system was developed by combining color co-occurrence matrix based texture analysis and g-ratio algorithms in C++ to acquire and process images in real-time to differentiate mowed wild blueberry plants from bare spots and weeds. The performance accuracy of the spot-applicable fertilizer spreader was evaluated both in laboratory simulation and real-time field tests. Simulation results reported that the accuracy of the developed system was 94.9 %. Real-time field tests reported that the system produced acceptable results at ground speeds of 1.6 and 3.2 km h−1 for the spot-application of fertilizer at target areas (in plant areas only) within the field. Results also indicated that the ground speed of 4.8 km h−1 was unacceptable, which could be due to blurred images at high speed and surface unevenness of the wild blueberry field. Spot-application of fertilizer using the modified fertilizer spreader could save fertilizer for the wild blueberry producers.

Wild blueberry (Vaccinium angustifolium Ait.) fields in the north east Canada are naturally grown in a course textured thin layer of soil and below this layer is a soilless layer of gravel. The root zone depth of this crop varies from 10 to 15 cm. Investigating the depth to the gravel layer below the course textured soil is advantageous, as it affects the water holding capacity of the root zone. Water and nutrient management are the two primary determinants of crop yield and the amount of leaching. The objective of this study was to estimate the depth to the gravel layer using DualEM-2 instrument. A C++ program written in Visual Studio 2010 was used to develop mathematical models for estimating the depth to the gravel layer from the outputs of DualEM-2 sensor. Two wild blueberry fields were selected in central Nova Scotia, Canada to evaluate the performance of DualEM-2 instrument in estimating the rootzone depth above the gravel layer. The mid points of squares created by grid lines were used as the sampling points at each experimental site. The actual depth to the interface was measured manually at selected grid points (n = 50). The apparent ground conductivity (ECa) values of DualEM-2 were recorded and the depth to the interface was estimated for the same sampling points within the selected fields. The fruit yield samples were also collected from the same grid points to identify the impact of the depth to the gravel layer on crop yield. After calibrations, comprehensive surveys were conducted and the actual and estimated depths to the interface were established. The interpolated maps of fruit yield, and the actual (zin) and estimated ( $$ {\text{z}}_{\text{in}}^{*} $$ ) depths to the interface were created in ArcGIS 10 software. Results indicated that the zin was significantly correlated with $$ {\text{z}}_{\text{in}}^{*} $$ for the North River (R 2 = 0.73; RMSE = 0.27 m) and the Carmel (R 2 = 0.45; RMSE = 0.20 m) sites. Results revealed that the areas with shallow depth to the gravel layer were low yielding, indicating that the variation in the depth to the gravel layer can have an impact on crop productivity. Non-destructive estimations of the depth to the gravel layer can be used to develop erosion control strategies, which will result in an increased crop production.

Wild blueberry fields have a significant proportion of bare spots/weed patches, scattered throughout the field, emphasizing the need for targeted fertilizer applications. The existing prescription map-based variable rate (VR) spreader cannot take account of the presence of small, irregularly shaped bare spots/weed patches during fertilization as only each half of the boom (3.66 m; 6 nozzles) is controlled—not each individual nozzle. The objective of this study was to modify the existing VR spreader for spot-application (SA) of fertilizer only in plant areas. The sensing and control system developed was capable of discriminating bare spots/weeds from plants, and shifting the independent control from 6 nozzles (3.66 m) to each pair of nozzles (1.22 m). The modified VR granular (MVRG) fertilizer spreader was able to use a prescription map and the sensing and control systems simultaneously within a field, to avoid fertilization in variable sized bare spots/weed patches. The performance of the MVRG fertilizer spreader was evaluated under two different lighting conditions and three ground speeds. Twelve bare spots/weed patches and plant areas were randomly selected in the wild blueberry field and weighed sticky collectors were placed to collect clay filler which was used as an analogue of fertilizer. The MVRG fertilizer spreader was operated on SA (application only in plant areas) and uniform (application in plants and bare spots/weed patches) modes alternately and the collectors were collected and re-weighed. Results of this study suggested that the MVRG spreader was capable of detecting plants/bare spots/weeds for SA of fertilizer. Spot application of fertilizer can help to reduce excessive fertilizer usage in bare spots/weeds, which will increase farm profitability.

Wild blueberry growers apply herbicides uniformly to control weeds within fields. The repeated and excessive use of herbicides in bare spots and blueberry plant areas that exist within fields has resulted in an increased cost of production and environmental contamination. A prototype variable rate (VR) sprayer was developed for spot-application (SA) of herbicides to the weeds. The boom was divided into 16 sections (0.76 m per section). The VR control system consisted of eight digital color cameras mounted on a separate boom in front of the tractor, a 20-channel MidTech Legacy 6000 controller, and two 8-channel VR controllers (VRCs) interfaced to a Pocket PC (PPC) via wireless Bluetooth®. Cameras were connected via USB cables to a computer. The cameras were capable of taking the images in real-time. Custom software was developed for processing the images to detect weeds in real-time. The triggering signals were sent to the VRC to open the specific nozzles where the weeds had been detected. The VR sprayer performed well and adequately sprayed propyzamide on red sorrel (Rumex acetosella L.) and flumioxazin on moss (Polytrichum commune) within the selected wild blueberry fields. Results of T-test from experiment 1 suggested that the percent area coverage of water sensitive papers in target and non-target areas after the flumioxazin application was significant indicating SA technique using VR sprayer accurately applied the flumioxazin on targets. SA from experiment 2 resulted in a 69% reduction in propyzamide use with 25 cm before and after target buffer and 28% cover of red sorrel.

An automated sensing system consisting of hardware and software was developed for spot-application of herbicide in pruned wild blueberry fields. The hardware of the system consisted of a ruggedized PC, four digital cameras, a signal output unit, an 8-channel computerized variable rate controller (VRC), a flow controller, a pocket PC, solenoid valves, a servo valve, a flow meter, and nozzles mounted on an all-terrain vehicle. Custom software was developed in C++ and installed in the ruggedized PC to acquire images from the cameras and process in real-time to calculate the fraction of green pixels for weed detection in each image. Green contrast between weed and non-weed area in pruned wild blueberry fields was used to develop a robust and effective discriminating algorithm. Red-Green-Blue (RGB) images were converted to normalized green ratio binary images using the ratio of 255*G/(R+G+B) followed by segmentation for the discrimination. Custom software was capable of processing images to differentiate weed area from non-weed area in real-time and send the signal to the VRC to open the valve in the specific section of the boom where the weeds were detected. The normalized green ratio algorithm showed very reliable accuracy under different outdoor light conditions (>500 lux) and shade conditions. The sensing system performed well for accurate weed detection and for sending signals to the VRC for spraying the correct weed targets at up to 8 km h-1 ground speed in the fields.

Wild blueberry producers apply agrochemicals uniformly without considering substantial variation in soil properties and topographic features that may affect fruit yield within a field and pollute environment. A wild blueberry field was selected in central Nova Scotia, Canada to evaluate the impact of variable rate (VR) fertilization on nutrient losses in surface runoff, from steep slope to low lying areas, and crop productivity. The field was divided into three subfields (Field 1, Field 2, and Field 3) under three management zones (Zone 1, Z1; Zone 2, Z2; and Zone 3, Z3). The runoff plots were installed to collect surface runoff samples from different zones of the field. Three fertilizer (NPK: 16.5 – 34.5 – 4.5, respectively) rates (200,150, and 100 kg ha -1 for Z1, Z2, Z3, respectively) were applied in runoff plots of VR treatment using fertilizer spreader. Uniform fertilizer rate of 200 kg ha -1 for comparison was applied in runoff plots of uniform application (UA) treatment. No fertilizer was applied in runoff plots of control treatment. The surface runoff samples were collected after every heavy rainfall event and analyzed for total phosphorus (TP), dissolved reactive phosphorus (DRP), particulate phosphorus (PP), and inorganic nitrogen concentrations. The VR treatment significantly (p≤0.05) decreased TP, DRP, PP, and inorganic nitrogen losses in surface runoff as compared to UA. Plant growth parameters and fruit yield were also collected from runoff plots. Plants in UA treatment were taller than VR treatment in low lying areas. Numbers of buds per stem were significantly affected by fertilizer treatments. The fruit yield in VR treatment was non-significantly different than UA treatment. The VR fertilizer application used 40% less fertilizer than standard UA of fertilizer. The VR fertilization could reduce nutrients loading in surface runoff from wild blueberry fields. The runoff plots could be used to accurately determine the surface runoff from small-scale runoff plots. However, the construction of runoff plots and data collection at field scale is laborious and time consuming.

The present study assessed the ability of a low cost dual-view computer-vision system (CVS) to measure volume and color co-occurrence image textural features of apple slices to find the end of drying process by comparing physical texture parameters and moisture content. Apple slices were dried at three different temperatures (40, 60 and 80 °C) and their volumes were measured with both the CVS and a caliper. The physical texture parameters and image textural features were measured according to the moisture content. While the CVS can be used to measure the volume of the apple slices during the drying process, the volume was not a good indicator of the end of drying because of porosity development. The peak force and initial slope of physical texture showed a significant correlation with H3, H5 and S2 and H3, H5 and S11 of image texture features, respectively. Eleven image texture features correlate well with the moisture content ( R 2 > 0.9) and also had a high correlation between each other. The uniformity of intensity (I1) of image feature was a good indicator of the end of drying because of the close correlations with moisture content. The drying times according to the moisture content and the uniformity of intensity were 4.31 ± 0.313 h and 4.42 ± 0.125 h, respectively.

Sensing and control system was developed for spot-application of fertilizer in real-time.Software was made to process data to differentiate foliage and bare spots.The modified fertilizer spreader was efficient and accurate enough for spot-application of fertilizer.Spot-application can increase profitability and mitigate the environmental risks. The commercial variable rate (VR) fertilizer spreader can change fertilizer rates in different management zones using global positioning system (GPS) guided prescription maps. The VR fertilizer spreader allows the independent control only on half of the fertilizer boom (3.66m; 6 nozzles). This existing VR fertilizer spreader fertilizes unevenly distributed small sized bare spots/weed patches which are less than half boom length in wild blueberry fields. Therefore, the existing VR fertilizer spreader was modified to control each pair of nozzles for spot-application of fertilizer only in plant areas. The automated sensing and control system consisting of 6µEye color cameras, solenoid valves, pneumatic cylinders, programmable logic controller (PLC), 8-channel VR controller (VRC) interfaced to a pocket PC (PPC) using wireless Bluetooth? radio with Windows Mobile? and a custom software was developed. The µEye color cameras were mounted on a separate boom in front of the tractor at a height of 1.2m. The modified VR spreader was capable of using prescription maps and automated sensing and control system simultaneously to detect foliage/bare spots in real-time to avoid fertilization in bare spots/weed patches. The ACCU-RATE controller calibration and response time calculations were performed prior to field experiment.The results of calibration tests for the ACCU-RATE controller suggested 0.05) between VR and UN treatments in grass/plant areas. These results suggested that the MVRG fertilizer spreader was efficient and accurate for spot-application of fertilizer to increase farm profitability and reduce environmental risks.

Wild blueberry producers apply fungicide uniformly without considering significant bare spots in the field. The wrong or over use of fungicide in bare spots results in an increased cost of production and threatens the environment. An automated prototype variable rate (VR) sprayer was used for spot-application (SA) of Chlorothalonil (Bravo®) fungicide in a wild blueberry field. Eighteen 6.1 m wide test tracks were selected in the field and bare spots were mapped using a real-time kinematics-global positioning system (RTK-GPS). Six plots were selected randomly for three different application rates. Water sensitive papers (WSP) were placed in foliage and bare spots in SA and uniform-application (UA) tracks. The percent area coverage (PAC) of WSP with both SA and UA in foliage and bare spot areas were calculated. Plant growth parameters were measured from all 108 randomly selected plots in SA, UA and control (CN) tracks for comparison. Plant images were taken over six selected plots in each of the 18 tracks. Images were analyzed using custom developed software to calculate the percentage of green pixels (PGP) for determining the effect of Bravo® on plant health. Fruit yield parameters were also measured from selected plots for comparison. Non-significance of the t test for SA versus UA plant targets’ PAC indicated that there was no significant bias in the SA with saving (9.90–51.22 %) and SA was accurate. Bravo® did not show any significant difference on plant growth parameters among SA, UA and CN. However, PGP, floral bud and harvestable yield of SA and UA were significantly increased over CN. Therefore, a VR sprayer could be used for SA of fungicides in wild blueberry cropping system to reduce chemical usage and maintain crop productivity.

Site-specific crop management can increase profitability and reduce environmental risks of wild blueberry crop having large spatial variation in soil/plant characteristics and topographic features which may affect fruit yield. An integrated automated system comprising of an ultrasonic sensor, a digital color camera, a slope sensor, a real-time kinematics GPS (RTK-GPS), custom software and ruggedized computer was developed. The system was incorporated onto a commercial wild blueberry harvester, to measure plant height, fruit yield, slope and elevation simultaneously while harvesting. Two wild blueberry fields were selected in central Nova Scotia, Canada to evaluate the performance of the system. Field boundaries, bare spots, weeds and grass patches were mapped with a RTK-GPS prior to start the experiment. Linear regression was used to calibrate the actual fruit yield with the percentage of blue pixels (R^2=0.91-0.92; P

Wild blueberry ( Vaccinium angustifolium Ait.) is a perennial rhizomatous low shrub and mostly mechanically harvested. The operator of the harvester needs to maintain the optimum height of harvester‘s head according to the plant height for better yield and quality while decreasing plant pulling. An ultrasonic on-the-go plant height measurement system (UPHMS II) was developed and compared with previous height measurement system (UPHMS I). A real-time kinematics differential global positioning system (RTK-GPS), a custom program and a ruggedized computer and both plant height sensing system were mounted on a commercial mechanical harvester for real-time plant height measurement during harvesting. A custom program was developed to acquire and process ultrasonic sensing data in real-time from both UPHMS I and UPHMS II simultaneously. Two wild blueberry fields were selected in central Nova Scotia to evaluate the performance of both UPHMS I and UPHMS II. Manually measured plant height values from 24 plots were compared with real-time measured values of two systems. UPHMS II performed better to predict wild blueberry plant height with higher accuracy than UPHMS I. The RMSE of the sensed height of UPHMS I and UPHMS II were 6.4 cm and 1.6 cm, respectively. UPHMS II can be an economic option to control wild blueberry harvester head automatically to increase harvester and operator‘s efficiency. Real-time and accurate sensing of plant height is the first step toward the automation of the wild blueberry harvester. Refinements for the optimum head height according to the plant height is required for future studies.

Similar to most biological studies, beef contamination classification studies using artificial neural networks are restricted to small datasets. This study evaluates multivariate normal (MVN) technique of synthetic sample generation on small datasets associated with Salmonella contamination in beef. Six experiments were conducted to evaluate the performance of integrated sensor system towards identification of Salmonella contaminated beef packages. Pattern recognition involved using wavelet packet transform for feature extraction from sensor array responses and radial basis function network (RBFN) based classification of contaminated beef packages from uncontaminated packages. The MVN generated synthetic olfactory sensor signatures were used to train and test the RBFN classifiers. For the datasets analyzed in this study, genetic algorithm optimized RBF networks conferred average contamination test classification accuracies of 90.33% ± 7.68% (mean ± std. dev.) which were higher compared to the bootstrapped quadratic discriminant analysis based average accuracies. RBFN classifier based average overall classification accuracies of six synthetically generated datasets were in the range of 86.66%–98.89% with highest average overall classification accuracies of 98.89% ± 1.92%.

Wild blueberry fruit yield maps could be used to generate prescription maps for site-specific application of fertilizer to improve crop productivity. An automated yield monitoring system II (AYMS II) consisting of two @mEye color cameras, real time kinematics-global positioning system, custom software, and a ruggedized laptop computer was developed and mounted on a Specialized Farm Motorized Vehicle (SFMV) for real-time fruit yield mapping. Custom software was developed in C++ programming language to acquire and process image in real-time from both cameras and store the percentage of blue pixels in ruggedized laptop computer. Two wild blueberry fields were selected in central Nova Scotia to evaluate the performance of the AYMS II for fruit yield mapping. Calibration was carried out at 20 randomly selected data points in each field. The ripe fruit was hand-harvested out of a 2.0x0.7m quadrat at each selected point and camera images were also taken from the same points to calculate the percentage of blue pixels (fraction of blue pixels in the image). Linear regression analysis was performed to calibrate the actual fruit yield with the percentage of blue pixels. Real-time yield mapping was carried out with AYMS II. The estimated yield along with geo-referenced coordinates was imported into ArcGIS 9.3 software for detailed mapping. The AYMS II, hardware and software performed well to estimate wild blueberry fruit yield. The linear regression results showed highly significant relationship between the percentage of blue pixels and actual fruit yield in field 1 and field 2. The correlation between actual and predicted fruit yield (validation, using the equation from field 2) in field 1 and field 2 (validation, using the equation from field 1) was also highly significant. Maps developed in ArcGIS 9.3 showed substantial variability in fruit yield in both fields. The bare spots were coincided with no or low yielding areas in the fields. The viable AYMS II will be incorporated into wild blueberry harvester. This information obtained from AYMS II could be used to implement site-specific management practices within the blueberry fields to optimize productivity while minimizing the environmental impact of farming operations.

Co-occurrence matrix-based textural features were analyzed and three algorithms were developed to identify bare spots, wild blueberry plants, and weeds with the aim of applying agrochemicals to wild blueberry cropping fields in a spot-specific manner. Images were acquired using four cameras and a ruggedized laptop with custom-written programs coded in Microsoft Visual® C++. Textural features were extracted from the images using MATLAB® and analyzed with SAS®. Forty-four textural features were extracted from co-occurrence matrices of NTSC luminance (L), hue, saturation, and intensity (HSI) images. Multiple discriminant analysis using all 44 features (DF_ALL model) showed 98.1% of overall classification accuracy and 83 ms of processing time of an image with C++ calculation. Based on the results of multiple discriminant analysis and two-step linear discrimination plotting, the DF_HSISD, DF_SISD, and HSILD algorithms are preferred algorithms with overall accuracy of 94.9%, 92.7%, and 91.4%, and processing time of 55, 27, and 29 ms, respectively. Any of three reduced textural feature algorithms can be employed for spot-specific application of agrochemicals in wild blueberry cropping fields. The choice of one algorithm over another will depend on whether processing speed or accuracy is more important for the end-user’s application.

An automated prototype variable rate (VR) sprayer was developed for control of 8 individual nozzles on a 6.1m sprayer boom for in-season, site-specific application of agrochemicals on weeds. The sprayer boom was divided into 8 sections and mounted behind an all-terrain vehicle (ATV) at 76.2cm above the ground. The variable-rate control system consisted of 8 ultrasonic sensors (one per spray section) mounted on a separate boom in front of the ATV, DICKEY-john Land Manager II controller and flow valve, solenoid valves and an 8-channel variable rate controller interfaced to a Pocket PC (PPC) using wireless Bluetooth^(R) radio with Windows Mobile^(R) compatible software. This type of VR sprayer does not use prescription maps, but relies on sensors to provide real-time weed detection information which is used to dispense correct agrochemical rates for the weeds. The sprayer can be used for in-season, spot application (SA) of agrochemicals by activating specific boom sections where the weeds have been detected. Two wild blueberry fields have been selected in central Nova Scotia to evaluate the accuracy of the VR sprayer. Water sensitive papers (targets) were stapled to weeds randomly selected in two tracks of each field. The papers were orientated parallel to the ground. The percent area coverage (PAC) of the sprayed targets with both SA and uniform application was calculated by an imaging system. Non significance of the t-test for uniform versus SA targets PAC indicated that there was no significant bias in the SA and that the SA was accurate. The PAC with both applications ranged from 10.01% to 81.22% and from 5.39% to 72.67% in field 1 and field 2, respectively. Weed heights at selected points were measured and related with percent area coverage to examine the sprayer performance for spray application. Linear regression analysis showed that weed heights were significantly correlated (R^2 ranged from 0.60 to 0.75) with percent area coverage of the targets in selected fields with both applications. It is proposed that herbicide should be applied at early stage of weed growth (weed height ranged from 35cm to 55cm and plant height ranged from 12cm to 30cm) for appropriate application with these specific VR sprayer arrangements. Based on these results, the VR sprayer was efficient and accurate enough for spot-application of agrochemicals usage in wild blueberry fields.

Wild blueberries are one of the most important fruit crops of Canada, producing more than 50% of the world’s production. Understanding and predicting the relationships between the machine operating parameters, fruit losses, topographic features and crop characteristics can aid in better berry recovery during mechanical harvesting. This paper suggested a modeling approach for prediction of fruit losses during harvesting using artificial neural network (ANN) and multiple regression (MR) techniques. Four wild blueberry fields were selected and completely randomized factorial (3 x 3) experiments were constructed at each site. One hundred sixty two plots (0.91 x 3 m) were made at each site, in the path of operating harvester. The total fruit yield, total losses were collected from each plot within selected fields. The harvester was operated at specific levels of ground speed (1.20, 1.60 and 2.00 km h-1) and head rotational speed (26, 28 and 30 rpm). The readings of slope, plant height, and fruit zone were also recorded from each plot. The collected data were normalized, and 70% of the data were utilized for training, and 30% for validation of the developed models using ANN and MR techniques. The developed models were validated internally and externally and the best performing model was identified based on mean square error (MSE), root mean square error (RMSE), coefficient of efficiency (CE) and coefficient of determination (R2). Results of scatter plot among the RMSE and epoch suggested that an epoch size of 15000 was appropriate to process fruit losses using ANN approach. Results revealed that the prediction accuracy of the MR models was lower (R2 = 0.46; RMSE = 0.14) than the ANN model (R2 = 0.84; RMSE = 0.075) for training dataset, which might be due to the non-linear nature of the data. Results reported that the ANN model predicted fruit losses with higher (R2 = 0.63; RMSE = 0.11) accuracy when compared with MR model (R2 = 0.37; RMSE = 0.15) for external validation dataset. Overall, the results of the study suggested that the ANN model was able to predict fruit losses accurately and reliably as functions of fruit yield, crop and machine variables. This study will help to identify the factors responsible for fruit losses and to suggest optimal harvesting scenarios to improve berry picking efficiency and recovery.

This study was designed to evaluate the performance efficiency of the commercial wild blueberry harvester for fruit losses during harvesting. Two wild blueberry fields were selected in central Nova Scotia, Canada. A 3 x 3 factorial experiment was constructed to examine the joint effect of ground speed and header revolution per minute (rpm) on picking efficiency of the harvester. Eighty one yield plots (0.91 x 3 m) were selected randomly in each field. The field boundaries, bare spots, weeds and yield plots were mapped with a real-time kinematics global positioning system (RTK-GPS). The harvester was operated at specific levels of ground speed (1.20, 1.6 and 2.0 km h -1 ) and header rpm (26, 28 and 30) to collect total fruit yield, un-harvested berries on the plants, berries on the ground, and berries through blower, from each plot within selected fields. The treatment combinations of the chosen factors were assigned randomly within selected fields. The slope, plant height, and fruit zone were also recorded manually from each plot. The results of classical statistics suggested that the pre-harvest fruit losses were lower in early season of harvesting. Un-harvested berries on the plants and losses through blower were significantly lower than losses on the ground for selected fields. Significant relationship between the fruit yield and total losses suggested that losses during harvesting were linear function of fruit yield within selected fields. Factorial analysis of variance (ANOVA) showed that ground speed, header rpm and their interaction were found to have significant impact on the picking efficiency of the harvester. Results of means comparison showed that a combination of 1.2 km h -1 and 26 rpm resulted in significantly lower losses when compared with other treatment combinations. Results also revealed that an ideal combination of ground speed and header rpm can minimize the fruit losses during harvesting, which will increase harvestable yield and ultimately enhance farm profitability.

Existing commercial variable rate granular (VRG) fertilizer spreader change fertilizer rates in different management zones using global positioning system-guided prescription maps without considering unevenly distributed bare spots and weed patches in wild blueberry fields. The modifications in the existing spreader were made to control each pair of nozzles independently to avoid excessive fertilization in bare spots and weed patches. For this purpose, six I¼Eye digital color cameras were incorporated on the boom (6.21 m wide) in front of the tractor. Custom software installed in the computer was capable of processing the images to detect weeds, bare spots and blueberry plants in real-time and sending signals to the 8-channel variable rate controller (VRC). The VRC then sends the signal to solenoid valves, which in turn closes the pair of nozzles which is directly above the bare spot or weed patch. This study was designed to evaluate the modified VRG (MVRG) fertilizer spreader for spot application of fertilizer only in plant areas. A preliminary test using clay filler as the applied material was conducted at the selected field and 7.3 m wide and 60 m long track was made in the field. Twelve artificial bare spots were made in the selected track using orange colored tarps. Equal numbers of weighed rat catchers (i.e. 12) were placed on artificial bare spots and grass. The MVRG fertilizer spreader was operated on variable rate (VR) mode (detection and no application of clay filler on bare spots). The rat catchers were collected and re-weighed. The same procedure was repeated for uniform (UN) application mode of the system. The results of paired t-test showed significant difference (p-value

Wild blueberry growers apply agrochemicals uniformly to control weeds within fields. The repeated and excessive use of agrochemicals in bare spots that exist within fields and on plants has resulted in increased cost of production and polluted environment. A commercial prototype variable rate (VR) sprayer was developed for spot-application of agrochemicals in a specific section of the 12.2 m sprayer boom where the weeds have been detected. The boom was divided into 16 sections (97 cm each section). VR control system consisted of eight digital color cameras mounted on a separate boom in front of the tractor, 20-channel MidTech Legacy 6000 controller and two 8-channel VR controllers interfaced to a Pocket PC using wireless Bluetooth® radio. Cameras were attached using USB serial cables to the computer. Custom software was capable of processing the images to detect weeds in real-time, and weed triggering signals were sent to the VRC to spray in the specific boom section where the weeds have been detected.

Wild blueberry yields are highly dependent on fungicides to control floral blight (monilinia and botrytis) and leaf diseases (septoria and rust). Growers apply fungicides uniformly without considering the significant bare spots (30-50% of the total field area). The repeated and excessive use of agrochemicals in bare spots has resulted in an increased cost of production. The over use of the fungicides is harmful to the environment and can contaminate surface and ground water. The proper targeting of fungicide on only foliage has the potential to save a substantial amount of chemical. Therefore, a significant need of an affordable variable rate sprayer for spot application of fungicide is needed for wild blueberry production. The objective of this study was to determine the performance of a prototype variable rate sprayer for spot application of Bravo® fungicide in wild blueberry fields. The 6.1 meter VR sprayer was mounted on a all-terrain vehicle. Color cameras were used to detect bare soil areas in wild blueberry fields. Eighteen 6.1 meter wide test plots were selected in a wild blueberry field and the bare soil areas were mapped using RTK-DGPS. Three application rates (uniform, variable and control) were selected at random for each plot. Digital color images were taken at 6 randomly selected locations in each of the 18 plots. Each image was analyzed to calculate green ratio for determining effect of Bravo® on wild blueberries. The application tracks were statistically compared with reference to the control tracks. Water sensitive paper was also placed in randomly selected locations for analysis purposes. The results can be used to determine the performance of applying fungicide on site-specific bases using a variable rate sprayer.

A novel metal oxide based electronic nose system was designed and developed to detect Salmonella contamination in packaged meat (beef). This system contained an array of metal oxide detectors and custom made electronics for acquiring olfactory signature of the volatile organic compounds in the headspace of the meat packages produced by detector arrays. The acquired signals were processed using three wavelet packet transforms (WPT) for simultaneous noise reduction and compression. The coefficients of wavelet packets were used as features for developing statistical classification models using bootstrapped linear and quadratic discriminant analysis techniques. The average total classification accuracies for classifying Salmonella contaminated meat samples were higher than 82% with highest average accuracy of about 88%.

Food safety is a critical need to ensure public safety. Application of olfactory sensing techniques for identifying the volatile metabolites produced by microbial pathogens is a potential method for detection of food contamination. A multidisciplinary effort using parallel sensing techniques is underway at North Dakota State University to detect Salmonella contamination in packaged meat. One of the focuses of the group is to develop and evaluate novel sensor/detector that will have high sensitivity and specificity for specific indicator compounds (trace levels). In this regard, zinc oxide semi-conductor material was selected as a novel sensor material to develop sensors. The objective of the present study was to fabricate and evaluate nanoparticulate ZnO sensor for the detection of lower concentration of ethanol and acetic acid at low operating temperature. The characterization of electrodes using atomic force microscopy, scanning electron microscopy and x-ray diffraction revealed the formation of nanoparticulate structures with (100), (101) and (002) orientation of zinc oxide. The preliminary results on gas sensitivity of ZnO indicated that the electrode displays a response to acetic acid, while they did not display any sensitivity to ethanol. When sensitivity of ZnO electrodes to 50 ppm and 100 ppm acetic acid were measured at 50oC, the electrodes displayed unstable responses. It is believed that high operating temperature may be required to improve electrode sensitivity to gases.