A study of 2D features for 3D visual SLAM

Paper submitted to the 43rd International Symposium on Robotics (ISR), Taipei, Taiwan, August 29-31, 2012.

Resolution limits to object tracking with subpixel accuracy

Subpixel methods increase the accuracy and efficiency of image detectors, processing units, and algorithms and provide very cost-effective systems for object tracking. Published methods achieve resolution increases up to three orders of magnitude. In this Letter, we demonstrate that this limit can be theoretically improved by several orders of magnitude, permitting micropixel and submicropixel accuracies. The necessary condition for movement detection is that one single pixel changes its status. We show that an appropriate target design increases the probability of a pixel change for arbitrarily small shifts, thus increasing the detection accuracy of a tracking system. The proposal does not impose severe restriction on the target nor on the sensor, thus allowing easy experimental implementation.

Improving 3D Keypoint Detection from Noisy Data Using Growing Neural Gas

3D sensors provides valuable information for mobile robotic tasks like scene classification or object recognition, but these sensors often produce noisy data that makes impossible applying classical keypoint detection and feature extraction techniques. Therefore, noise removal and downsampling have become essential steps in 3D data processing. In this work, we propose the use of a 3D filtering and down-sampling technique based on a Growing Neural Gas (GNG) network. GNG method is able to deal with outliers presents in the input data. These features allows to represent 3D spaces, obtaining an induced Delaunay Triangulation of the input space. Experiments show how the state-of-the-art keypoint detectors improve their performance using GNG output representation as input data. Descriptors extracted on improved keypoints perform better matching in robotics applications as 3D scene registration.

Use of Costas arrays in subpixel metrology

Subpixel methods increase the accuracy and efficiency of image detectors, processing units, and algorithms and provide very cost-effective systems for object tracking. A recently proposed method permits micropixel and submicropixel accuracies providing certain design constraints on the target are met. In this paper, we explore the use of Costas arrays - permutation matrices with ideal auto-ambiguity properties - for the design of such targets.

Conformational changes of dissolved humic and fulvic superstructures with progressive iron complexation

Conformational changes of a humic acid (HA) and a fulvic acid (FA) induced by iron complexation were followed by high-performance size exclusion chromatography (HPSEC) with both UV–vis and refractive index (RI) detectors. Molecular size distribution was reduced for HA and increased for FA with progressive iron complexation. Since interactions of Fe with humic components are electrostatic, it is likely that the triple-charged Fe ions formed stronger complexes with the more acidic hydrophilic and hydrated FA than with the less acidic and more hydrophobic HA. The large content of ionized carboxyl groups in FA, thus favored the formation of intra- or intermolecular bridges between the negatively charged fulvic acid molecules, and led to more compact and larger size network than for HA. Conversely, iron complexation with HA disrupted the humic conformational arrangements stabilized by only weak hydrophobic bonds into smaller-size aggregates of greater conformational stability due to formation of strong metal complexes. These results confirmed that humic molecules in solution were organized in supramolecular associations of relatively small molecules loosely bound together by dispersive interactions and hydrogen bonds, and they specifically responded to chemical changes brought about by metal additions. The present study revealed the molecular changes occurring in superstructures of natural organic matter when in metal complexes and contributed to understand and predict the environmental behavior in waters and soil of metal complexes with natural organic matter.

An improvement of a SLAM RGB-D method with movement prediction derived from a study of visual features

This paper presents a method for the fast calculation of a robot’s egomotion using visual features. The method is part of a complete system for automatic map building and Simultaneous Location and Mapping (SLAM). The method uses optical flow to determine whether the robot has undergone a movement. If so, some visual features that do not satisfy several criteria are deleted, and then egomotion is calculated. Thus, the proposed method improves the efficiency of the whole process because not all the data is processed. We use a state-of-the-art algorithm (TORO) to rectify the map and solve the SLAM problem. Additionally, a study of different visual detectors and descriptors has been conducted to identify which of them are more suitable for the SLAM problem. Finally, a navigation method is described using the map obtained from the SLAM solution.

Semantic localization in the PCL library

The semantic localization problem in robotics consists in determining the place where a robot is located by means of semantic categories. The problem is usually addressed as a supervised classification process, where input data correspond to robot perceptions while classes to semantic categories, like kitchen or corridor. In this paper we propose a framework, implemented in the PCL library, which provides a set of valuable tools to easily develop and evaluate semantic localization systems. The implementation includes the generation of 3D global descriptors following a Bag-of-Words approach. This allows the generation of fixed-dimensionality descriptors from any type of keypoint detector and feature extractor combinations. The framework has been designed, structured and implemented to be easily extended with different keypoint detectors, feature extractors as well as classification models. The proposed framework has also been used to evaluate the performance of a set of already implemented descriptors, when used as input for a specific semantic localization system. The obtained results are discussed paying special attention to the internal parameters of the BoW descriptor generation process. Moreover, we also review the combination of some keypoint detectors with different 3D descriptor generation techniques.

On the Electric Breakdown in Liquid Argon at Centimeter Scale

We present a study on the dependence of electric breakdown discharge properties on electrode geometry and the breakdown field in liquid argon near its boiling point. The measurements were performed with a spherical cathode and a planar anode at distances ranging from 0.1 mm to 10.0 mm. A detailed study of the time evolution of the breakdown volt-ampere characteristics was performed for the first time. It revealed a slow streamer development phase in the discharge. The results of a spectroscopic study of the visible light emission of the breakdowns complement the measurements. The light emission from the initial phase of the discharge is attributed to electro-luminescence of liquid argon following a current of drifting electrons. These results contribute to set benchmarks for breakdown-safe design of ionization detectors, such as Liquid Argon Time Projection Chambers (LAr TPC).

Accelerator and detector physics at the Bern medical cyclotron and its beam transport line.

The cyclotron laboratory for radioisotope production and multi-disciplinary research at the Bern University Hospital (Inselspital) is based on an 18-MeV proton accelerator, equipped with a specifically conceived 6-m long external beam line, ending in a separate bunker. This facility allows performing daily positron emission tomography (PET) radioisotope production and research activities running in parallel. Some of the latest developments on accelerator and detector physics are reported. They encompass novel detectors for beam monitoring and studies of low current beams.

Detection of low energy antimatter with emulsions

Emulsion detectors feature a very high position resolution and consequently represent an ideal device when particle detection is required at the micrometric scale. This is the case of quantum interferometry studies with antimatter, where micrometric fringes have to be measured. In this framework, we designed and realized a new emulsion based detector characterized by a gel enriched in terms of silver bromide crystal contents poured on a glass plate. We tested the sensitivity of such a detector to low energy positrons in the range 10–20 keV . The obtained results prove that nuclear emulsions are highly efficient at detecting positrons at these energies. This achievement paves the way to perform matter-wave interferometry with positrons using this technology.

Dosco Machine

IMB (Irvine, Michigan, Brookline), a collaboration between the University of Michigan, the University of California at Irvine, and the U.S. Department of Energy, was an experiment designed to determine the ultimate stability of matter. Construction image. Mechanical "mole" (dosco machine) brought in by Morton Salt Co. to dig the cavity of IMB, 1979-1980

Ramp into cavity

IMB (Irvine, Michigan, Brookline), a collaboration between the University of Michigan, the University of California at Irvine, and the U.S. Department of Energy, was an experiment designed to determine the ultimate stability of matter. Construction image. Finished cavity -- looking through tunnel to main part of mine.