No abstract

Given the problem that the existing method of station distributing the pseudosatellite system cannot ensure both its coverage and position in a situation of signal occlusion, it proposed a new stationary layout method with an elite strategy for a ground-based pseudosatellite positioning system based on the elite strategy of the nondominant genetic rankings (NSGA-II). The geometrical design of the pseudosatellite system is calculated by visual domain analysis and precision factors for the signal coverage age and base station. To optimize the algorithm, the NSGA-II algorithm is used. An earth pseudosatellite positioning system method of stationary distribution is obtained that simultaneously optimizes signal coverage and positioning accuracy. The algorithm is better distributed and has a certain superintendence compared with the traditional genetic algorithm.

No abstract

Most high-performance semantic segmentation networks are based on complicated deep convolutional neural networks, leading to severe latency in real-time detection. However, the state-of-the-art semantic segmentation networks with low complexity are still far from detecting objects accurately. In this paper, we propose a real-time semantic segmentation network, RecepNet, which balances accuracy and inference speed well. Our network adopts a bilateral architecture (including a detail path, a semantic path and a bilateral aggregation module). We devise a lightweight baseline network for the semantic path to gather rich semantic and spatial information. We also propose a detail stage pattern to store optimized high-resolution information after removing redundancy. Meanwhile, the effective feature-extraction structures are designed to reduce computational complexity. RecepNet achieves an accuracy of 78.65% mIoU (mean intersection over union) on the Cityscapes dataset in the multi-scale crop and flip evaluation. Its algorithm complexity is 52.12 GMACs (giga multiply–accumulate operations) and its inference speed on an RTX 3090 GPU is 50.12 fps. Moreover, we successfully applied RecepNet for blue-green algae real-time detection. We made and published a dataset consisting of aerial images of water surface with blue-green algae, on which RecepNet achieved 82.12% mIoU. To the best of our knowledge, our dataset is the world’s first public dataset of blue-green algae for semantic segmentation.

It is a crucial premise for Fizeau interferometric imaging that all the optical axes of sub-apertures are well aligned. What is challenging in tip/tilt alignment of such interferometers is the superposition of all the sub-aperture spots on the focal plane for generation of fringes. In this paper, we propose and demonstrate a simple tip/tilt alignment scheme via defocus-based sub-spot separation (TADS). The focal interference function is divided into sparse defocused sub-spots at a long defocusing distance. Then the servo-control technology with a centroid extracting algorithm is used to align all the sub-aperture optical axes, each corresponding to a defocused sub-spot. The pointing correction performance using TADS has been verified through closed-loop experiments performed on a Fizeau telescope array testbed newly built in our laboratory. Compared with the current means, the proposed TADS eliminates the requirement of complex hardware design and the limitation of sub-aperture amounts, which is compatible with most existing segmented mirrors and telescope arrays.