Developed data fusion method of remotely sensed imagery data (high-resolution satellite and UAS multispectral imagery) to improve phenomics protocols in pulse crop breeding programsBuilding machine learning-based disability weight to measure disease burden on dairy cattle using pathophysiological and behavioral data Created an image-based classifier using deep learning (CNN) to classify lentil root images into disease severity categories, leading to successfully classifying the disease-resistant roots with an F1 score of 90% Automated chromatography-based peak analysis method to assess volatile organic compounds for disease resistance characterization and detection

Evaluated different remote sensing imagery data (i.e. optical and synthetic aperture radar SAR data) for monitoring regenerative practices in rice; extracted imagery features using traditional computer vision techniques and built a machine learning model to classify these practices

Collaborated with a cross-departmental team to quantify root rot disease resistance in more than 500 lentil lines/accessions and identified a set of digital traits (RGB and hyperspectral) using regression LASSO, leading to accurately detecting genes associated with disease resistance Designed and optimized low-cost imaging settings to screen more than 10,000 samples of lentil rootsDeveloped a multispectral image processing algorithm for canopy cover estimation in wheat breeding trials and implemented k-means clustering for background segmentation