arXiv subjects

Robert Brunner

Publications and source records attributed to Robert Brunner.

13 recordsLinked to original sources

arXiv2020

The 1st Agriculture-Vision Challenge: Methods and Results

The first Agriculture-Vision Challenge aims to encourage research in developing novel and effective algorithms for agricultural pattern recognition from aerial images, especially for the semantic segmentation task associated with our challenge dataset. Around 57 participating teams from various countries compete to achieve state-of-the-art in aerial agriculture semantic segmentation. The Agriculture-Vision Challenge Dataset was employed, which comprises of 21,061 aerial and multi-spectral farmland images. This paper provides a summary of notable methods and results in the challenge. Our submission server and leaderboard will continue to open for researchers that are interested in this challenge dataset and task; the link can be found here.

cs.CV↗

arXiv2020

Agriculture-Vision: A Large Aerial Image Database for Agricultural Pattern Analysis

The success of deep learning in visual recognition tasks has driven advancements in multiple fields of research. Particularly, increasing attention has been drawn towards its application in agriculture. Nevertheless, while visual pattern recognition on farmlands carries enormous economic values, little progress has been made to merge computer vision and crop sciences due to the lack of suitable agricultural image datasets. Meanwhile, problems in agriculture also pose new challenges in computer vision. For example, semantic segmentation of aerial farmland images requires inference over extremely large-size images with extreme annotation sparsity. These challenges are not present in most of the common object datasets, and we show that they are more challenging than many other aerial image datasets. To encourage research in computer vision for agriculture, we present Agriculture-Vision: a large-scale aerial farmland image dataset for semantic segmentation of agricultural patterns. We collected 94,986 high-quality aerial images from 3,432 farmlands across the US, where each image consists of RGB and Near-infrared (NIR) channels with resolution as high as 10 cm per pixel. We annotate nine types of field anomaly patterns that are most important to farmers. As a pilot study of aerial agricultural semantic segmentation, we perform comprehensive experiments using popular semantic segmentation models; we also propose an effective model designed for aerial agricultural pattern recognition. Our experiments demonstrate several challenges Agriculture-Vision poses to both the computer vision and agriculture communities. Future versions of this dataset will include even more aerial images, anomaly patterns and image channels. More information at https://www.agriculture-vision.com.

cs.CV↗

arXiv2017

Workload Analysis of Blue Waters

Blue Waters is a Petascale-level supercomputer whose mission is to enable the national scientific and research community to solve "grand challenge" problems that are orders of magnitude more complex than can be carried out on other high performance computing systems. Given the important and unique role that Blue Waters plays in the U.S. research portfolio, it is important to have a detailed understanding of its workload in order to guide performance optimization both at the software and system configuration level as well as inform architectural balance tradeoffs. Furthermore, understanding the computing requirements of the Blue Water's workload (memory access, IO, communication, etc.), which is comprised of some of the most computationally demanding scientific problems, will help drive changes in future computing architectures, especially at the leading edge. With this objective in mind, the project team carried out a detailed workload analysis of Blue Waters.

cs.DC↗

arXiv2012

Bring out your codes! Bring out your codes! (Increasing Software Visibility and Re-use)

Progress is being made in code discoverability and preservation, but as discussed at ADASS XXI, many codes still remain hidden from public view. With the Astrophysics Source Code Library (ASCL) now indexed by the SAO/NASA Astrophysics Data System (ADS), the introduction of a new journal, Astronomy & Computing, focused on astrophysics software, and the increasing success of education efforts such as Software Carpentry and SciCoder, the community has the opportunity to set a higher standard for its science by encouraging the release of software for examination and possible reuse. We assembled representatives of the community to present issues inhibiting code release and sought suggestions for tackling these factors. The session began with brief statements by panelists; the floor was then opened for discussion and ideas. Comments covered a diverse range of related topics and points of view, with apparent support for the propositions that algorithms should be readily available, code used to produce published scientific results should be made available, and there should be discovery mechanisms to allow these to be found easily. With increased use of resources such as GitHub (for code availability), ASCL (for code discovery), and a stated strong preference from the new journal Astronomy & Computing for code release, we expect to see additional progress over the next few years.

astro-ph.IM↗

arXiv2012

The SDSS DR7 Galaxy Angular Power Spectrum: Volume-Limits and Galaxy Morphology

We use a quadratic estimator with KL-compression to calculate the angular power spectrum of a volume-limited Sloan Digital Sky Survey (SDSS) Data Release 7 (DR7) galaxy sample out to l = 200. We also determine the angular power spectrum of selected subsamples with photometric redshifts z < 0.3 and 0.3 < z < 0.4 to examine the possible evolution of the angular power spectrum, as well as early-type and late-type galaxy subsamples to examine the relative linear bias. In addition, we calculate the angular power spectrum of the SDSS DR7 main galaxy sample in a ~ 53.7 square degree area out to l = 1600 to determine the SDSS DR7 angular power spectrum to high multipoles. We perform a \chi^2 fit to compare the resulting angular power spectra to theoretical nonlinear angular power spectra to extract cosmological parameters and the linear bias. We find the best-fit cosmological parameters of \Omega_m = 0.267 +- 0.038 and \Omega_b = 0.045 +- 0.012. We find an overall linear bias of b = 1.075 +- 0.056, an early-type bias of b_e = 1.727 +- 0.065, and a late-type bias of b_l = 1.256 +- 0.051. Finally, we present evidence of a selective misclassification of late-type galaxies as stars by the SDSS photometric data reduction pipeline in areas of high stellar density (e.g., at low Galactic latitudes).

astro-ph.CO↗

arXiv2011

The SDSS DR7 Galaxy Angular Power Spectrum

We calculate the angular power spectrum of galaxies selected from the Sloan Digital Sky Survey (SDSS) Data Release 7 (DR7) by using a quadratic estimation method with KL-compression. The primary data sample includes over 18 million galaxies covering more than 5,700 square degrees after masking areas with bright objects, reddening greater than 0.2 magnitudes, and seeing of more than 1.5 arcseconds. We test for systematic effects by calculating the angular power spectrum by SDSS stripe and find that these measurements are minimally affected by seeing and reddening. We calculate the angular power spectrum for l \leq 200 multipoles by using 40 bandpowers for the full sample, and l \leq 1000 multipoles using 50 bandpowers for individual stripes. We also calculate the angular power spectrum for this sample separated into 3 magnitude bins with mean redshifts of z = 0.171, z = 0.217, and z = 0.261 to examine the evolution of the angular power spectrum. We determine the theoretical linear angular power spectrum by projecting the 3D power spectrum to two dimensions for a basic comparison to our observational results. By minimizing the {\chi}^2 fit between these data and the theoretical linear angular power spectrum we measure a loosely-constrained fit of {\Omega}_m = 0.31^{+0.18}_{-0.11} with a linear bias of b = 0.94 \pm 0.04.

astro-ph.CO↗

arXiv2007

Angular Power Spectrum Estimation using High Performance Reconfigurable Computing

Angular power spectra are an important measure of the angular clustering of a given distribution. In Cosmology, they are applied to such vastly different observations as galaxy surveys that cover a fraction of the sky and the Cosmic Microwave Background that covers the entire sky, to obtain fundamental parameters that determine the structure and evolution of the universe. The calculation of an angular power spectrum, however, is complex and the optimization of these calculations is a necessary consideration for current and forthcoming observational surveys. In this work, we present preliminary results of implementing angular power spectrum estimation scheme on a high-performance reconfigurable computing platform.

astro-ph↗

arXiv2005

The DEEP Groth Strip Survey. I. The Sample

The Deep Extragalactic Exploratory Probe (DEEP) is a multi-phase research program dedicated to the study of the formation and evolution of galaxies and of large scale structure in the distant Universe. This paper describes the first five-year phase, denoted DEEP1. A series of ten DEEP1 papers will discuss a range of scientific topics (e.g., the study of photometric and spectral properties of a general distant galaxy survey, the evolution observed in galaxy populations of varied morphologies). The observational basis for these studies is the Groth Survey Strip field, a 127 square arcminute region which has been observed with the Hubble Space Telescope in both broad I-band and V-band optical filters and with the Low Resolution Imaging Spectrograph on the Keck Telescopes. Catalogs of photometric and structural parameters have been constructed for 11,547 galaxies and stars at magnitudes brighter than 29, and spectroscopy has been conducted for a magnitude-color weighted subsample of 818 objects. We evaluate three independent techniques for constructing an imaging catalog for the field from the HST data, and discuss the depth and sampling of the resultant catalogs. The selection of the spectroscopic subsample is discussed, and we describe the multifaceted approach taken to prioritizing objects of interest for a variety of scientific subprograms. A series of Monte Carlo simulations then demonstrates that the spectroscopic subsample can be adequately modeled as a simple function of magnitude and color cuts in the imaging catalog.

astro-ph↗

arXiv2005

An Empirical Calibration of the Completeness of the SDSS Quasar Survey

Spectra of nearly 20000 point-like objects to a Galactic reddening corrected magnitude of i=19.1 have been obtained to test the completeness of the SDSS quasar survey. The spatially-unresolved objects were selected from all regions of color space, sparsely sampled from within a 278 sq. deg. area of sky covered by this study. Only ten quasars were identified that were not targeted as candidates by the SDSS quasar survey (including both color and radio source selection). The inferred density of unresolved quasars on the sky that are missed by the SDSS algorithm is 0.44 per sq. deg, compared to 8.28 per sq. deg. for the selected quasar density, giving a completeness of 94.9(+2.6,-3.8) to the limiting magnitude. Omitting radio selection reduces the color-only selection completeness by about 1%. Of the ten newly identified quasars, three have detected broad absorption line systems, six are significantly redder than other quasars at the same redshift, and four have redshifts between 2.7 and 3.0 (the redshift range where the SDSS colors of quasars intersect the stellar locus). The fraction of quasars missed due to image defects and blends is approximately 4%, but this number varies by a few percent with magnitude. Quasars with extended images comprise about 6% of the SDSS sample, and the completeness of the selection algorithm for extended quasars is approximately 81%, based on the SDSS galaxy survey. The combined end-to-end completeness for the SDSS quasar survey is approximately 89%. The total corrected density of quasars on the sky to i=19.1 is estimated to be 10.2 per sq. deg.

astro-ph↗

arXiv2003

A Snapshot Survey for Gravitational Lenses Among z>=4.0 Quasars: I. The z>5.7 Sample

Over the last few years, the Sloan Digital Sky Survey (SDSS) has discovered several hundred quasars with redshift between 4.0 and 6.4. Including the effects of magnification bias, one expects a priori that an appreciable fraction of these objects are gravitationally lensed. We have used the Advanced Camera for Surveys on the Hubble Space Telescope to carry out a snapshot imaging survey of high-redshift SDSS quasars to search for gravitationally split lenses. This paper, the first in a series reporting the results of the survey, describes snapshot observations of four quasars at z = 5.74, 5.82, 5.99 and 6.30, respectively. We find that none of these objects has a lensed companion within 5 magnitudes with a separation larger than 0.3 arcseconds; within 2.5 magnitudes, we can rule out companions within 0.1 arcseconds. Based on the non-detection of strong lensing in these four systems, we constrain the z~6 luminosity function to a slope of beta>-4.63 (3 sigma), assuming a break in the quasar luminosity function at M_{1450}^*=-24.0. We discuss the implications of this constraint on the ionizing background due to quasars in the early universe. Given that these quasars are not highly magnified, estimates of the masses of their central engines by the Eddington argument must be taken seriously, possibly challenging models of black hole formation.

astro-ph↗

arXiv2001

Weak Lensing Measurements of 42 SDSS/RASS Galaxy Clusters

We present a lensing study of 42 galaxy clusters imaged in Sloan Digital Sky Survey (SDSS) commissioning data. Cluster candidates are selected optically from SDSS imaging data and confirmed for this study by matching to X-ray sources found independently in the ROSAT all sky survey (RASS). Five color SDSS photometry is used to make accurate photometric redshift estimates that are used to rescale and combine the lensing measurements. The mean shear from these clusters is detected to 2 h-1 Mpc at the 7-sigma level, corresponding to a mass within that radius of 4.2 +/- 0.6 x 10^14 h-1 M_sun. The shear profile is well fit by a power law with index -0.9 +/- 0.3, consistent with that of an isothermal density profile. This paper demonstrates our ability to measure ensemble cluster masses from SDSS imaging data.

astro-ph↗

arXiv2000

Discovery of a Close Pair of z = 4.25 Quasars from the Sloan Digital Sky Survey

We report the discovery of a pair of z = 4.25 quasars with a separation of 33 arcseconds. The brighter of the two objects was identified as a high-redshift quasar candidate from Sloan Digital Sky Survey multicolor imaging data, and the redshift was measured from a spectrum obtained with the Hobby-Eberly Telescope. The slit orientation of this observation {\it by chance} included another quasar, approximately one magnitude fainter and having the same redshift as the target. This is the third serendipitous discovery of a z > 4 quasar. The differences in the relative strengths and profiles of the emission lines suggest that this is a quasar pair and not a gravitational lens. The two objects are likely to be physically associated; the projected physical separation is approximately 210 $h_{50}^{-1}$ kpc and the redshifts are identical to $\approx$ 0.01, implying a radial physical separation of 950 $h_{50}^{-1}$ kpc or less. The existence of this pair is strong circumstantial evidence that $z \sim 4$ quasars are clustered.

astro-ph↗

arXiv1997

A2125 and its Environs: Evidence for an X-ray-emitting Hierarchical Superstructure

Based on a deep ROSAT/PSPC observation, we reveal an elongated complex of extended X-ray-emitting objects in and around the galaxy cluster A2125. Multicolor optical imaging of galaxies in the field suggests that this complex represents a hierarchical superstructure spanning about 11 Mpc at redshift z = 0.247. The multi-peak X-ray morphology of A2125 suggests that the cluster is an ongoing coalescence of at least three major subunits. The dynamical youth of this cluster is consistent with its large fraction of blue galaxies observed by Butcher & Oemler. The superstructure contains two additional clusters, projected at distances of only 3 and 4.3 Mpc from A2125. But the most interesting feature is the low-surface-brightness X-ray emission from a moderate galaxy concentration not associated with individual clusters. The emission likely arises in a hot intergalactic medium, as predicted in N-body/hydro simulations of structure formation.

astro-ph↗