Year-end Sale % OFF 
Abstract
Unmanned aerial vehicle (UAV) imaging is a promising data acquisition technique for image-based plant phenotyping. However, UAV images have a lower spatial resolution than similarly equipped in field ground-based vehicle systems, such as carts, because of their distance from the crop canopy, which can be particularly problematic for measuring small-sized plant features. In this study, the performance of three deep learning-based super resolution models, employed as a pre-processing tool to enhance the spatial resolution of low resolution images of three different kinds of crops were evaluated. To train a super resolution model, aerial images employing two separate sensors co-mounted on a UAV flown over lentil, wheat and canola breeding trials were collected. A software workflow to pre-process and align real-world low resolution and high-resolution images and use them as inputs and targets for training super resolution models was created. To demonstrate the effectiveness of real-world images, three different experiments employing synthetic images, manually downsampled high resolution images, or real-world low resolution images as input to the models were conducted. The performance of the super resolution models demonstrates that the models trained with synthetic images cannot generalize to real-world images and fail to reproduce comparable images with the targets. However, the same models trained with real-world datasets can reconstruct higher-fidelity outputs, which are better suited for measuring plant phenotypes.
Highlights
The demand for sustainable food production continues to increase due to population growth, in the developing world [1]
A paired LR/HR image dataset of aerial images does not currently exist, and is more challenging to collect, as it requires capturing both LR and HR images of a dynamic scene. We address these shortcomings by providing a paired drone image set of low and high resolution images, and characterize the performance against state of the art super resolution algorithms to establish the utility of super resolution approaches in agriculture, and more generally for other applications from drone images
We evaluate the performance of the super resolution models trained and tested with each individual experiment
Summary
The demand for sustainable food production continues to increase due to population growth, in the developing world [1]. Image-based plant phenotyping has the potential to increase the speed and reliability of phenotyping by using remote sensing technology, such as unmanned aerial vehicles (UAVs), to estimate phenomic information from field crop breeding experiments [7,8,9,10,11]. There are usually a variety of ground objects in the sensor’s field of view, where a pixel in a remote sensing image may contain a mixture of ground objects information, called mixed pixel. This can impact information extraction, such as image classification and object detection in mixed pixels [38]. A finer spatial resolution results in smaller pixel sizes and fewer mixed pixels in an image [41]
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
