VTechWorks staff will be away for the winter holidays starting Tuesday, December 24, 2024, through Wednesday, January 1, 2025, and will not be replying to requests during this time. Thank you for your patience, and happy holidays!

Browsing by Author "Robbins, James"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Applications of High-Resolution Imaging for Open Field Container Nursery Counting
    She, Ying; Ehsani, Reza; Robbins, James; Nahún Leiva, Josué; Owen, James S. Jr. (MDPI, 2018-12-12)
    Frequent inventory data of container nurseries is needed by growers to ensure proper management and marketing strategies. In this paper, inventory data are estimated from aerial images. Since there are thousands of nursery species, it is difficult to find a generic classification algorithm for all cases. In this paper, the development of classification methods was confined to three representative categories: green foliage, yellow foliage, and flowering plants. Vegetation index thresholding and the support vector machine (SVM) were used for classification. Classification accuracies greater than 97% were obtained for each case. Based on the classification results, an algorithm based on canopy area mapping was built for counting. The effects of flight altitude, container spacing, and ground cover type were evaluated. Results showed that container spacing and interaction of container spacing with ground cover type have a significant effect on counting accuracy. To mimic the practical shipping and moving process, incomplete blocks with different voids were created. Results showed that the more plants removed from the block, the higher the accuracy. The developed algorithm was tested on irregular- or regular-shaped plants and plants with and without flowers to test the stability of the algorithm, and accuracies greater than 94% were obtained.
  • Thumbnail Image
    The Relationship between Drone Speed and the Number of Flights in RFID Tag Reading for Plant Inventory
    Quino, Jannette; Maja, Joe Mari; Robbins, James; Owen, James; Chappell, Matthew; Camargo, Joao Neto; Fernandez, R. Thomas (MDPI, 2021-12-22)
    Accurate inventory allows for more precise forecasting, including profit projections, easier monitoring, shorter outages, and fewer delivery interruptions. Moreover, the long hours of physical labor involved over such a broad area and the effect of inefficiencies could lead to less accurate inventory. Unreliable data and predictions, unannounced stoppages in operations, production delays and delivery, and a considerable loss of profit can all arise from inaccurate inventory. This paper extends our previous work with drones and RFID by evaluating: the number of flights needed to read all tags deployed in the field, the number of scans per pass, and the optimum drone speed for reading tags. The drone flight plan was divided into eight passes from southwest to northwest and back at a horizontal speed of 2.2, 1.7, and 1.1 m per second (m/s) at a vertically fixed altitude. The results showed that speed did not affect the number of new tags scanned (p-value > 0.05). Results showed that 90% of the tags were scanned in less than four trips (eight passes) at 1.7 m/s. Based on these results, the system can be used for large-scale nursery inventory and other industries that use RFID tags in outdoor environments. We presented two novel measurements on evaluating RFID reader efficiency by measuring how fast the reader can read and the shortest distance traveled by the RFID reader over tag.
  • Thumbnail Image
    Watson on the Farm: Using Cloud-Based Artificial Intelligence to Identify Early Indicators of Water Stress
    Freeman, Daniel; Gupta, Shaurya; Smith, D. Hudson; Maja, Joe Mari; Robbins, James; Owen, James S.; Peña, Jose M.; de Castro, Ana I. (MDPI, 2019-11-13)
    As demand for freshwater increases while supply remains stagnant, the critical need for sustainable water use in agriculture has led the EPA Strategic Plan to call for new technologies that can optimize water allocation in real-time. This work assesses the use of cloud-based artificial intelligence to detect early indicators of water stress across six container-grown ornamental shrub species. Near-infrared images were previously collected with modified Canon and MAPIR Survey II cameras deployed via a small unmanned aircraft system (sUAS) at an altitude of 30 meters. Cropped images of plants in no, low-, and high-water stress conditions were split into four-fold cross-validation sets and used to train models through IBM Watson’s Visual Recognition service. Despite constraints such as small sample size (36 plants, 150 images) and low image resolution (150 pixels by 150 pixels per plant), Watson generated models were able to detect indicators of stress after 48 hours of water deprivation with a significant to marginally significant degree of separation in four out of five species tested (p < 0.10). Two models were also able to detect indicators of water stress after only 24 hours, with models trained on images of as few as eight water-stressed Buddleia plants achieving an average area under the curve (AUC) of 0.9884 across four folds. Ease of pre-processing, minimal amount of training data required, and outsourced computation make cloud-based artificial intelligence services such as IBM Watson Visual Recognition an attractive tool for agriculture analytics. Cloud-based artificial intelligence can be combined with technologies such as sUAS and spectral imaging to help crop producers identify deficient irrigation strategies and intervene before crop value is diminished. When brought to scale, frameworks such as these can drive responsive irrigation systems that monitor crop status in real-time and maximize sustainable water use.