Synchronization of two independently moving cameras without feature correspondences

In this work, a method that synchronizes two video sequences is proposed. Unlike previous methods, which require the existence of correspondences between features tracked in the two sequences, and/or that the cameras are static or jointly moving, the proposed approach does not impose any of these constraints. It works when the cameras move independently, even if different features are tracked in the two sequences. The assumptions underlying the proposed strategy are that the intrinsic parameters of the cameras are known and that two rigid objects, with independent motions on the scene, are visible in both sequences. The relative motion between these objects is used as clue for the synchronization. The extrinsic parameters of the cameras are assumed to be unknown. A new synchronization algorithm for static or jointly moving cameras that see (possibly) different parts of a common rigidly moving object is also proposed. Proof-of-concept experiments that illustrate the performance of these methods are presented, as well as a comparison with a state-of-the-art approach.

Stress-induced cortisol secretion impairs detection performance in x-ray baggage screening for hidden weapons by screening novices

Aviation security strongly depends on screeners' performance in the detection of threat objects in x-ray images of passenger bags. We examined for the first time the effects of stress and stress-induced cortisol increases on detection performance of hidden weapons in an x-ray baggage screening task. We randomly assigned 48 participants either to a stress or a nonstress group. The stress group was exposed to a standardized psychosocial stress test (TSST). Before and after stress/nonstress, participants had to detect threat objects in a computer-based object recognition test (X-ray ORT). We repeatedly measured salivary cortisol and X-ray ORT performance before and after stress/nonstress. Cortisol increases in reaction to psychosocial stress induction but not to nonstress independently impaired x-ray detection performance. Our results suggest that stress-induced cortisol increases at peak reactivity impair x-ray screening performance.

Fully automatic segmentation of AP pelvis X-rays via random forest regression with efficient feature selection and hierarchical sparse shape composition

In clinical practice, traditional X-ray radiography is widely used, and knowledge of landmarks and contours in anteroposterior (AP) pelvis X-rays is invaluable for computer aided diagnosis, hip surgery planning and image-guided interventions. This paper presents a fully automatic approach for landmark detection and shape segmentation of both pelvis and femur in conventional AP X-ray images. Our approach is based on the framework of landmark detection via Random Forest (RF) regression and shape regularization via hierarchical sparse shape composition. We propose a visual feature FL-HoG (Flexible- Level Histogram of Oriented Gradients) and a feature selection algorithm based on trace radio optimization to improve the robustness and the efficacy of RF-based landmark detection. The landmark detection result is then used in a hierarchical sparse shape composition framework for shape regularization. Finally, the extracted shape contour is fine-tuned by a post-processing step based on low level image features. The experimental results demonstrate that our feature selection algorithm reduces the feature dimension in a factor of 40 and improves both training and test efficiency. Further experiments conducted on 436 clinical AP pelvis X-rays show that our approach achieves an average point-to-curve error around 1.2 mm for femur and 1.9 mm for pelvis.

A new method to measure higher visual functions in an immersive environment

BACKGROUND: Higher visual functions can be defined as cognitive processes responsible for object recognition, color and shape perception, and motion detection. People with impaired higher visual functions after unilateral brain lesion are often tested with paper pencil tests, but such tests do not assess the degree of interaction between the healthy brain hemisphere and the impaired one. Hence, visual functions are not tested separately in the contralesional and ipsilesional visual hemifields. METHODS: A new measurement setup, that involves real-time comparisons of shape and size of objects, orientation of lines, speed and direction of moving patterns, in the right or left visual hemifield, has been developed. The setup was implemented in an immersive environment like a hemisphere to take into account the effects of peripheral and central vision, and eventual visual field losses. Due to the non-flat screen of the hemisphere, a distortion algorithm was needed to adapt the projected images to the surface. Several approaches were studied and, based on a comparison between projected images and original ones, the best one was used for the implementation of the test. Fifty-seven healthy volunteers were then tested in a pilot study. A Satisfaction Questionnaire was used to assess the usability of the new measurement setup. RESULTS: The results of the distortion algorithm showed a structural similarity between the warped images and the original ones higher than 97%. The results of the pilot study showed an accuracy in comparing images in the two visual hemifields of 0.18 visual degrees and 0.19 visual degrees for size and shape discrimination, respectively, 2.56° for line orientation, 0.33 visual degrees/s for speed perception and 7.41° for recognition of motion direction. The outcome of the Satisfaction Questionnaire showed a high acceptance of the battery by the participants. CONCLUSIONS: A new method to measure higher visual functions in an immersive environment was presented. The study focused on the usability of the developed battery rather than the performance at the visual tasks. A battery of five subtasks to study the perception of size, shape, orientation, speed and motion direction was developed. The test setup is now ready to be tested in neurological patients.

Dish Detection and Segmentation for Dietary Assessment on Smartphones

Diet-related chronic diseases severely affect personal and global health. However, managing or treating these diseases currently requires long training and high personal involvement to succeed. Computer vision systems could assist with the assessment of diet by detecting and recognizing different foods and their portions in images. We propose novel methods for detecting a dish in an image and segmenting its contents with and without user interaction. All methods were evaluated on a database of over 1600 manually annotated images. The dish detection scored an average of 99% accuracy with a .2s/image run time, while the automatic and semi-automatic dish segmentation methods reached average accuracies of 88% and 91% respectively, with an average run time of .5s/image, outperforming competing solutions.

Food Recognition for Dietary Assessment Using Deep Convolutional Neural Networks

Diet management is a key factor for the prevention and treatment of diet-related chronic diseases. Computer vision systems aim to provide automated food intake assessment using meal images. We propose a method for the recognition of already segmented food items in meal images. The method uses a 6-layer deep convolutional neural network to classify food image patches. For each food item, overlapping patches are extracted and classified and the class with the majority of votes is assigned to it. Experiments on a manually annotated dataset with 573 food items justified the choice of the involved components and proved the effectiveness of the proposed system yielding an overall accuracy of 84.9%.

A Closed-Form, Consistent and Robust Solution to Uncalibrated Photometric Stereo Via Local Diffuse Reflectance Maxima

Images of an object under different illumination are known to provide strong cues about the object surface. A mathematical formalization of how to recover the normal map of such a surface leads to the so-called uncalibrated photometric stereo problem. In the simplest instance, this problem can be reduced to the task of identifying only three parameters: the so-called generalized bas-relief (GBR) ambiguity. The challenge is to find additional general assumptions about the object, that identify these parameters uniquely. Current approaches are not consistent, i.e., they provide different solutions when run multiple times on the same data. To address this limitation, we propose exploiting local diffuse reflectance (LDR) maxima, i.e., points in the scene where the normal vector is parallel to the illumination direction (see Fig. 1). We demonstrate several noteworthy properties of these maxima: a closed-form solution, computational efficiency and GBR consistency. An LDR maximum yields a simple closed-form solution corresponding to a semi-circle in the GBR parameters space (see Fig. 2); because as few as two diffuse maxima in different images identify a unique solution, the identification of the GBR parameters can be achieved very efficiently; finally, the algorithm is consistent as it always returns the same solution given the same data. Our algorithm is also remarkably robust: It can obtain an accurate estimate of the GBR parameters even with extremely high levels of outliers in the detected maxima (up to 80 % of the observations). The method is validated on real data and achieves state-of-the-art results.

GoCARB: Smartphone-App zur Kohlenhydratberechnung bei Typ-1-Diabetes

Smartphone-App zur Kohlenhydratberechnung Neue Technologien wie Blutzuckersensoren und moderne Insulinpumpen prägten die Therapie des Typ-1-Diabetes (T1D) in den letzten Jahren in wesentlichem Ausmaß. Smartphones sind aufgrund ihrer rasanten technischen Entwicklung eine weitere Plattform für Applikationen zur Therapieunterstützung bei T1D. GoCARB Hierbei handelt es sich um ein zur Kohlenhydratberechnung entwickeltes System für Personen mit T1D. Die Basis für Endanwender stellt ein Smartphone mit Kamera dar. Zur Berechnung werden 2 mit dem Smartphone aus verschiedenen Winkeln aufgenommene Fotografien einer auf einem Teller angerichteten Mahlzeit benötigt. Zusätzlich ist eine neben dem Teller platzierte Referenzkarte erforderlich. Die Grundlage für die Kohlenhydratberechnung ist ein Computer-Vision-gestütztes Programm, das die Mahlzeiten aufgrund ihrer Farbe und Textur erkennt. Das Volumen der Mahlzeit wird mit Hilfe eines dreidimensional errechneten Modells bestimmt. Durch das Erkennen der Art der Mahlzeiten sowie deren Volumen kann GoCARB den Kohlenhydratanteil unter Einbeziehung von Nährwerttabellen berechnen. Für die Entwicklung des Systems wurde eine Bilddatenbank von mehr als 5000 Mahlzeiten erstellt und genutzt. Resümee Das GoCARB-System befindet sich aktuell in klinischer Evaluierung und ist noch nicht für Patienten verfügbar.

Lung Pattern Classification for Interstitial Lung Diseases Using a Deep Convolutional Neural Network

Automated tissue characterization is one of the most crucial components of a computer aided diagnosis (CAD) system for interstitial lung diseases (ILDs). Although much research has been conducted in this field, the problem remains challenging. Deep learning techniques have recently achieved impressive results in a variety of computer vision problems, raising expectations that they might be applied in other domains, such as medical image analysis. In this paper, we propose and evaluate a convolutional neural network (CNN), designed for the classification of ILD patterns. The proposed network consists of 5 convolutional layers with 2×2 kernels and LeakyReLU activations, followed by average pooling with size equal to the size of the final feature maps and three dense layers. The last dense layer has 7 outputs, equivalent to the classes considered: healthy, ground glass opacity (GGO), micronodules, consolidation, reticulation, honeycombing and a combination of GGO/reticulation. To train and evaluate the CNN, we used a dataset of 14696 image patches, derived by 120 CT scans from different scanners and hospitals. To the best of our knowledge, this is the first deep CNN designed for the specific problem. A comparative analysis proved the effectiveness of the proposed CNN against previous methods in a challenging dataset. The classification performance (~85.5%) demonstrated the potential of CNNs in analyzing lung patterns. Future work includes, extending the CNN to three-dimensional data provided by CT volume scans and integrating the proposed method into a CAD system that aims to provide differential diagnosis for ILDs as a supportive tool for radiologists.

Building the View Graph of a Category by Exploiting Image Realism

We propose a weakly supervised method to arrange images of a given category based on the relative pose between the camera and the object in the scene. Relative poses are points on a sphere centered at the object in a given canonical pose, which we call object viewpoints. Our method builds a graph on this sphere by assigning images with similar viewpoint to the same node and by connecting nodes if they are related by a small rotation. The key idea is to exploit a large unlabeled dataset to validate the likelihood of dominant 3D planes of the object geometry. A number of 3D plane hypotheses are evaluated by applying small 3D rotations to each hypothesis and by measuring how well the deformed images match other images in the dataset. Correct hypotheses will result in deformed images that correspond to plausible views of the object, and thus will likely match well other images in the same category. The identified 3D planes are then used to compute affinities between images related by a change of viewpoint. We then use the affinities to build a view graph via a greedy method and the maximum spanning tree.

Artists' advance: decreased upper alpha power while drawing in artists compared with non-artists

Brain mechanisms associated with artistic talents or skills are still not well understood. This exploratory study investigated differences in brain activity of artists and non-artists while drawing previously presented perspective line-drawings from memory and completing other drawing-related tasks. Electroencephalography (EEG) data were analyzed for power in the frequency domain by means of a Fast Fourier Transform (FFT). Low Resolution Brain Electromagnetic Tomography (LORETA) was applied to localize emerging significances. During drawing and related tasks, decreased power was seen in artists compared to non-artists mainly in upper alpha frequency ranges. Decreased alpha power is often associated with an increase in cognitive functioning and may reflect enhanced semantic memory performance and object recognition processes in artists. These assumptions are supported by the behavioral data assessed in this study and complement previous findings showing increased parietal activations in non-artists compared to artists while drawing. However, due to the exploratory nature of the analysis, additional confirmatory studies will be needed.

Automatic extraction of femur contours from calibrated fluoroscopic images

Automatic identification and extraction of bone contours from X-ray images is an essential first step task for further medical image analysis. In this paper we propose a 3D statistical model based framework for the proximal femur contour extraction from calibrated X-ray images. The automatic initialization is solved by an estimation of Bayesian network algorithm to fit a multiple component geometrical model to the X-ray data. The contour extraction is accomplished by a non-rigid 2D/3D registration between a 3D statistical model and the X-ray images, in which bone contours are extracted by a graphical model based Bayesian inference. Preliminary experiments on clinical data sets verified its validity