Jure Skvarč | Searcharxiv

arXiv2026

The WEAVE acquisition and guiding software: pattern recognition-based acquisition and multi-fibre guiding

We present the architecture, implementation, and on-sky validation of the fully automated acquisition and guiding system (AG) developed for the WEAVE instrument on the William Herschel Telescope. The AG operates in two distinct modes, corresponding to the observing modes of WEAVE. For the large integral field unit (LIFU), an off-axis imaging guider is used, for which we have devised an automatic acquisition method based on pattern recognition of stellar asterisms matched against Gaia predictions. For the multi-object spectrograph (MOS) and the mini-integral field units (mIFU), a multi-fibre guider uses up to eight coherent image guide fibre bundles to derive and apply continuous corrections in azimuth, altitude, and rotation. The system performs complete astrometric calculations, including atmospheric differential refraction and instrument flexure, for each guide frame, enabling accurate target placement and stable closed-loop guiding in all configurations. To support development, commissioning, and operational validation, we have also built a high-fidelity simulation mode that reproduces the behaviour of the telescope control system and of the AG cameras, and we release the standalone camera simulator as open-source software. Using two years of routine WEAVE operations spanning commissioning and early survey phases, we present a statistically robust characterization of AG performance, demonstrating that both modes meet design requirements and are ready for sustained survey operations.

astro-ph.IM↗

arXiv2019

Segmentation-Based Deep-Learning Approach for Surface-Defect Detection

Automated surface-anomaly detection using machine learning has become an interesting and promising area of research, with a very high and direct impact on the application domain of visual inspection. Deep-learning methods have become the most suitable approaches for this task. They allow the inspection system to learn to detect the surface anomaly by simply showing it a number of exemplar images. This paper presents a segmentation-based deep-learning architecture that is designed for the detection and segmentation of surface anomalies and is demonstrated on a specific domain of surface-crack detection. The design of the architecture enables the model to be trained using a small number of samples, which is an important requirement for practical applications. The proposed model is compared with the related deep-learning methods, including the state-of-the-art commercial software, showing that the proposed approach outperforms the related methods on the specific domain of surface-crack detection. The large number of experiments also shed light on the required precision of the annotation, the number of required training samples and on the required computational cost. Experiments are performed on a newly created dataset based on a real-world quality control case and demonstrates that the proposed approach is able to learn on a small number of defected surfaces, using only approximately 25-30 defective training samples, instead of hundreds or thousands, which is usually the case in deep-learning applications. This makes the deep-learning method practical for use in industry where the number of available defective samples is limited. The dataset is also made publicly available to encourage the development and evaluation of new methods for surface-defect detection.

cs.CV↗