Extended faint source detection in astronomical hyperspectral images

Abstract

International audienceThe circum-galactic medium consists in gas orbiting around galaxies, whose faintness prevents any complete and easy detection. A powerful tool to detect such pattern can be found in using hyperspectral imaging. Nevertheless, detection in hyperspectral datacubes faces various problems, including well-fitted signal and noise descriptions to ensure further discrimination. A specificity of astronomical images resides in dealing with faint and very noisy signals. In this paper, we introduce a new constrained generalized likelihood ratio test adapted to the problem and a compound test to exploit most of the available information. We also investigate the use of both spatial information and multiple observations on a single scene, to enhance robustness. Numerical experiments on synthetic data are performed to quantify the gain of the different approaches. Finally, results on real hyperspectral astronomical data are presented, which may map for the first time observation of the circum-galactic medium around faint and distant galaxie

Pairwise Markov fields for segmentation in astronomical hyperspectral images

“…The intensities are then convolved by a 3 × 3 Moffat FSF (cf. figures 2c, 2b and also [7]). Finally, µ contains 3 non-zero coefficients, as shown in Fig.…”

“…Fig. 5 reports the performance comparison of the MGCoPMF model with [7] and [2,1]. The investigation of these results reveals the following points:…”

“…Six methods are evaluated: 5. the detection method introduced in [7], based on an hypothesis-testing framework. We use the same experimental settings proposed as in [7] concerning the choice of the catalog and the P FA , and use the same FSF as in the synthetic images ;…”

“…The problem of faint source detection in astronomical HSI was recently addressed by dedicated methods within an hypothesis testing framework [7,2,1], in which the absence and presence of signal are two competing hypothesis. In [7], we accounted for spatial and spectral priors to perform a local contextual detection, and errors were handled locally with the expected Probability of False Alarm (PFA). In [2,1], the authors use a matched-filter preprocessing to tackle the spatial structure, and the errors are globally handled through robust False Discovery Rate (FDR) control.…”

“…This model is applied for the detection of signals in extremely noisy HSI (Section 2 and Section 3). The proposed model is evaluated on synthetic images, and it is shown that it outperforms the state-of-the-art methods from [7,2,1] shows that these latter are outperformed by the proposed method (Section 4). Finally, the segmentation of real MUSE images provide satisfying results with regards to the target application (Section 5).…”