Search

Currently, we can solve a wide range of tasks using computer vision algorithms, which reduce manual labor and enable rapid analysis of the environment. The remote sensing domain provides vast amounts of satellite data, but it also poses challenges associated with processing this data. Baseline solutions with intermediate results are available for various tasks, such as forest species classification, infrastructure recognition, and emergency situation analysis using satellite data. Despite these advances, two major issues with high-performing artificial intelligence algorithms remain in the current decade. The first issue relates to the availability of data. To train a robust algorithm, a reasonable amount of well-annotated training data is required. The second issue is the availability of satellite data, which is another concern. Even though there are a number of data providers, high-resolution and up-to-date imagery is extremely expensive. This paper aims to address these challenges by proposing an effective pipeline for building segmentation that utilizes freely available Sentinel-2 data with 10 m spatial resolution. The approach we use combines a super-resolution (SR) component with a semantic segmentation component. As a result, we simultaneously consider and analyze SR and building segmentation tasks to improve the quality of the infrastructure analysis through medium-resolution satellite data. Additionally, we collected and made available a unique dataset for the Russian Federation covering area of 1091.2 square kilometers. The dataset provides Sentinel-2 imagery adjusted to the spatial resolution of 2.5 m and is accompanied by semantic segmentation masks. The building footprints were created using OpenStreetMap data that was manually checked and verified. Several experiments were conducted for the SR task, using advanced image SR methods such as the diffusion-based SR3 model, RCAN, SRGAN, and MCGR. The MCGR network produced the best result, with a PSNR of 27.54 and SSIM of 0.79. The obtained SR images were then used to tackle the building segmentation task with different neural network models, including DeepLabV3 with different encoders, SWIN, and Twins transformers. The SWIN transformer achieved the best results, with an F1-score of 79.60. [ABSTRACT FROM AUTHOR]

[Display omitted] • Deep learning-based apple tree health assessment system using image segmentation. • End-to-end implementation with a user-friendly interface. • The developed system was tested under field conditions in an apple orchard. • The platform allows to speed up the digital phenotyping process. Computer vision and machine learning have recently been applied to a number of sensing platforms, boosting their performance to a new level. These advances have shown the vast possibilities for enhancing remote plant health assessment and disease detection. Until now, however, the scanning time and spatial resolution of such automated tools have been limited, as well as the area of application. We developed a state-of-the-art sensing system equipped with artificial intelligence and multispectral imaging with a special focus on near real-time and universality of application in agriculture. For this purpose, we collected a dataset of over 360,000 images of healthy and infected apple trees to develop and test our system, which includes a Convolutional Neural Network (CNN) algorithm for leaves segmentation. The proposed solution automatically computed vegetation indices (VIs) accurate to a single pixel. Further, we developed a desktop application for data post-processing and visualization, which allows the user to rapidly assess the health status of a vast agricultural area and thoroughly examine each tree individually. The developed system was successfully tested under field conditions in a large apple orchard, confirming viability of a reliable, end-to-end solution based on a computer vision platform for remote assessment of plant health and identification of stressed plants with high precision and spatial resolution. [ABSTRACT FROM AUTHOR]