Ongoing Projects
PhD Research Project
Processing and Analyzing Aerial as well as Ground Images of Different Crops in order to Estimate their Different Agricultural Phenotypes
In collaboration with Data Science Institute and School of Plant Sciences
As the header mentions, this is a joint project with the Data Science Institute and School of Plant Sciences at the University of Arizona. In this project, we are working with the data generated by the Gantry Machine at the research fields of the university. This machine takes numerous images from the field daily which further can be used to monitor the growth of individual crops. The final goal is to measure the phenotypic features of different variation and mutations of plants in different conditions, mainly under drought stress. At this stage of the project, we are processing the raw images to generate clean orthomosaics. We have developed a new robust method for doing such a thing. Another part of the project is identifying individual crops and monitor their growth through time. This part is done using Deep Neural Network. Another use of Neural Networks in this project is to segmentation of sorghum leaf blight disease on sorghum leaves. Find generated orthomosaics of lettuce at the given link. You can also take a look at the results of our stitching method on a full field orthomosaics of lettuce season at this link and compare it to an uncorrected mosaic at this link. Following illustrates some pictures related to the project. Hover your mouse over the images to read a short explanation about them. Download a short video of the machine at this link.
In collaboration with Data Science Institute and School of Plant Sciences
As the header mentions, this is a joint project with the Data Science Institute and School of Plant Sciences at the University of Arizona. In this project, we are working with the data generated by the Gantry Machine at the research fields of the university. This machine takes numerous images from the field daily which further can be used to monitor the growth of individual crops. The final goal is to measure the phenotypic features of different variation and mutations of plants in different conditions, mainly under drought stress. At this stage of the project, we are processing the raw images to generate clean orthomosaics. We have developed a new robust method for doing such a thing. Another part of the project is identifying individual crops and monitor their growth through time. This part is done using Deep Neural Network. Another use of Neural Networks in this project is to segmentation of sorghum leaf blight disease on sorghum leaves. Find generated orthomosaics of lettuce at the given link. You can also take a look at the results of our stitching method on a full field orthomosaics of lettuce season at this link and compare it to an uncorrected mosaic at this link. Following illustrates some pictures related to the project. Hover your mouse over the images to read a short explanation about them. Download a short video of the machine at this link.
PhD Research Project
Classification of Whole Slide Images (WSI) of Human Tissues into Different Disease classes Using Deep Neural Networks
In collaboration with Data Science Institute
In this project, we are employing Deep Neural Networks to classify patches of human breast tissue into normal and cancerous groups. Moreover, we have been trying to come up with a sound Neural Network to aggregate all these classifications and based on them, locat and detect the tumors in a Whole Slide Image (WSI). We have been focusing on making our method more interpretable by applying modifications on the network. In the following part, you can see some images with some explanations about them. Hover your mouse over the images to read the short explanations.
In collaboration with Data Science Institute
In this project, we are employing Deep Neural Networks to classify patches of human breast tissue into normal and cancerous groups. Moreover, we have been trying to come up with a sound Neural Network to aggregate all these classifications and based on them, locat and detect the tumors in a Whole Slide Image (WSI). We have been focusing on making our method more interpretable by applying modifications on the network. In the following part, you can see some images with some explanations about them. Hover your mouse over the images to read the short explanations.
