Jaroslaw Bulat

Data Processing Tasks in Wireless GI Endoscopy: Image-Based Capsule Localization & Navigation and Video Compression

The paper addresses data processing support that is required in capsule gastrointestinal endoscopy. First, capsule position estimation method using standard MPEG-7 image features (descriptors) is discussed. The proposed approach makes use of vector quantization, principal component analysis and neural networks. Next, new algorithms dedicated for virtual colonoscopy (VC) human body inspection are described. The VC images can be registered with endoscopic ones and help in capsule localization and navigation. Finally, an original, low- complexity, efficient image compression method, based on integer-to-integer 4x4 DCT transform, is presented and experimentally verified.

Reproducibility of crosstalk measurements on active glasses 3D LCD displays based on temporal characterization

Crosstalk is one of the main display-related perceptual factors degrading image quality and causing visual discomfort on 3D-displays. It causes visual artifacts such as ghosting effects, blurring, and lack of color fidelity which are considerably annoying and can lead to difficulties to fuse stereoscopic images. On stereoscopic LCD with shutter-glasses, crosstalk is mainly due to dynamic temporal aspects: imprecise target luminance (highly dependent on the combination of left-view and right-view pixel color values in disparity regions) and synchronization issues between shutter-glasses and LCD. These different factors influence largely the reproducibility of crosstalk measurements across laboratories and need to be evaluated in several different locations involving similar and differing conditions. In this paper we propose a fast and reproducible measurement procedure for crosstalk based on high-frequency temporal measurements of both display and shutter responses. It permits to fully characterize crosstalk for any right/left color combination and at any spatial position on the screen. Such a reliable objective crosstalk measurement method at several spatial positions is considered a mandatory prerequisite for evaluating the perceptual influence of crosstalk in further subjective studies.

Localization of Endoscopic Capsule in the GI Tract Based on MPEG-7 Visual Descriptors

The paper addresses the problem of localization of video endoscopic capsule in the gastrointestinal (GI) tract on the base of appropriate classification of images received from it. In this context usefulness of MPEG-7 image descriptors as classification features has been verified. For classification purpose various state of the art tools were used including Neural Networks and Vector Quantization. The dimension of the problem was also reduced by the Principal Component Analysis. Novelty of the presented approach consists in joint application of mentioned above techniques for recognition of the GI region inspected by the capsule by means of classification of MPEG-7 features to different parts of GI tract. In this research recognition of the upper part organs of the GI tract has been performed.

The BRONCHOVID – Computer System Supporting Bronchoscopy Laboratory

The bronchoscopy diagnostics consists of numerous tasks including but not limited to the treatment planning on the basis of computed tomography, registration and analysis of endoscopic images during the inspection of the tracheo-bronchial tree and procedure management. There is no comprehensive solution for a computer system supporting bronchoscopy laboratory. Since the limited time of medical personnel, high volume and multimodal data created during bronchoscopy procedure and complex workflow we decide to build a dedicated system addressing these issues. The paper presets an overview of the bronchoscope diagnostics, user needs and requirements related to the procedure and the proposed solution. Finely, the deployment results as well as future plan is presented.

Some new results from TEQ design for maximum bit rate ADSL modem

The paper presents examples of the successful application of a new mathematical SNR model to bit rate evaluation of asymmetric digital subscriber line (ADSL) modem. Contrary to other methods, the present design of a time domain equalizer (TEQ) additionally takes into account nonideal frequency response of demodulation filters. Also strong demodulation filters influence on achievable modem bit rate is shown. Different types of disturbances (such as AWGN, NEXT, narrow band interference) are considered. The verified model can be very useful during design and optimization of the TEQ filters for general multicarrier DMT/OFDM transceivers. As an example presented in the paper clearly shows, application of the new model in a cost function during TEQ filter optimization leads to an increase in the ADSL modem bit rate. It is also demonstrated that this improvement is not directly SSNR (SNR of the shortened impulse response of the channel) dependent.

Emerging Research Directions on 3D Video Quality Assessment

Motion picture producers, providers and equipment developers have to deeply consider end user perception of the application being often expressed in terms of a capacious Quality of Experience (QoE) concept. QoE is affected across the whole application delivery chain including content digitisation and compression, its network delivery and reproduction. During recent years enormous research effort and massive tests have been performed in order to identify factors affecting QoE and develop their mapping to scales like Mean Opinion Score for 2D content. Today, the digital video world is on the eve of 3D imaging which is far more complex and sophisticated not only because of the involved technology but also due to the multi–factor nature of the overall 3D experience. This paper discusses the current state of the research on the emerging problem of the user perceived quality of 3D content.

Computational tasks in computer-assisted transbronchial biopsy

The paper presents algorithmic solutions dedicated to computer navigation system which is to assist bronchoscope positioning during transbronchial needle-aspiration biopsy. The navigation exploits the principle of on-line registration of real images from endoscope camera and the virtual ones generated on the base of computed tomography (CT) data of a patient. When these images are similar an assumption is made, that the bronchoscope and virtual camera have approximately the same position and view direction. In this paper the following computational aspects are described: correction of camera lens distortion, fast approximate estimation of endoscope ego-motion, reconstruction of bronchial tree from the CT data by means of their segmentation and its centerline calculation, virtual views generation, registration of real and virtual images via maximization of their mutual information and, finally, efficient parallel and network implementation of the navigation system which is under development at present.

Computational Tasks in Bronchoscope Navigation During Computer-Assisted Transbronchial Biopsy

The paper presents algorithmic solutions dedicated to computer navigation system which is to assist bronchoscope positioning during transbronchial needle-aspiration biopsy. The navigation exploits principle of on-line registration of real images coming from endoscope camera and virtual ones generated on the base of computed-tomography (CT) data of a patient. When these images are similar an assumption is made that the bronchoscope and virtual camera have approximately the same position and view direction. In the paper the following computational aspects are described: correction of camera lens distortion, fast approximate estimation of endoscope ego-motion, reconstruction of bronchial tree from CT data by means of their segmentation and its centerline calculation, virtual views generation, registration of real and virtual images via maximization of their mutual information and, finally, efficient parallel and network implementation of the navigation system which is under development.

Minimum distortion TEQ equalizer design for DMT system

In this paper, the problem of time domain equalizer (TEQ) design for discrete multitone DMT applications is dealt with, based on a matrix time domain derivation of total distortion power of intersymbol interference (ISI) and noise in multicarrier modulation schemes with a cyclic prefix. We argue for a modification of the popular form of weighting shortened impulse response tail and head samples, contributing to ISI power. We stress the total power (not per-channel) nature of the solution, which means it is not directly bit rate oriented. However, the extension to a frequency domain bit rate maximizing approach has been sought as well and described in the cited reference. Despite all the numerical results, which are presented for the modification of one exemplary design algorithm presented recently, i.e. the eigenfilter of Vaidyanathan and Tkacenko, the main observation is valid for a wide range of design rules originating from the maximum shortening SNR approach.

Vision-based navigation assistance for visually impaired individuals using general purpose mobile devices

At the present there are some systems offering pedestrian navigation, but not as many are designated for people with visual impairments. The aim of this study is to present a navigation aid which makes use of a stereoscopy video system, OpenCV library, and general purpose mobile devices. The system is based on algorithms for machine image recognition using a stereoscopy camera rig. It reconstructs the surface of the surrounding scene with high accuracy, and is thus able to provide information about potential obstacles. Our experiments show it is possible to create a real-time solution using a contemporary Android-based mobile phone and two USB cameras.