3 June 2024 in IEEE Access
Modern greenhouses are characterized by scale, standardization, and efficiency. Some indoor agricultural facilities spread across over 100000 m 2 while being inhabited by dozens of species in a fragile balance. Timely monitoring is crucial for these greenhouses, and both speed and precision are required to minimize the potential yield loss. Human inspection is prone to error and time-consuming, thus expensive. On the other hand, collecting samples and analyzing them in the laboratory outside the agricultural facility is even more time-consuming and economically infeasible. Consequently, inspection by an autonomous robot that performs visual plant analysis gains traction. This work is devoted to the data acquisition and training of a neural network-based classifier that solves the problem of powdery mildew identification. The data collection was performed with user-grade RGB cameras while in motion. The infection identification accuracy of 85% was reached with a small computational load on the robot’s computer, which proves that the target problem could be solved under challenging circumstances. In order to carry out the experiments, an autonomous robot capable of greenhouse monitoring was developed and tested in a real environment. It is supposed to be deployed in modern greenhouses equipped with rails.
1 April 2025 in IEEE Access
Random sample consensus-based algorithm (RANSAC) is one of the standard approaches in the field of robust estimation of parametric objects. There are numerous versions of RANSAC that are tailored to specific problems such as homography estimation, point cloud registration, plane estimation and others. On the other hand, numerous attempts were made towards the development of modifications of RANSAC that are extremely robust to noise or highly efficient in terms of the computational demands. However, the general premise of these approaches remains all the same: instead of trying to fit a single model to all the data, a number of small models is produced and evaluated in terms of their ability to describe the data. In this work, an empirical study is carried out that is purposely limited to a specific application of RANSAC. Particularly, the dependency of the output quality on the noise is examined for the quadric surface evaluation. The datawas presented in the form of 3D point clouds. A method for synthetic data generation was developed, allowing to mimic the geometric principles of the point cloud formation. The dependency of the error on the noise was examined on the synthetic point clouds. Finally, numerical experiments on real point clouds were carried out in the agricultural setting. The code can be found in the repository https://github.com/aidagroup/SCUF/ and the proposed method can be used as a module https://pypi.org/project/scuf/.