Jaromir Przybylo

PixelStreams-based implementation of videodetector

This paper introduces a method that enhances the relocatability of partial bitstreams for FPGA run-time reconfiguration. Reconfigurable applications usually employ partial bitstreams which are specific to one target region on the FPGA. Previously, techniques have been proposed that allow relocation between identical regions on the FPGA. However, as FPGAs are becoming increasingly heterogeneous, this approach is often too restrictive. We introduce a method that circumvents the problem of having to find fully identical regions based on compatible subsets of resources, enabling flexible placement of relocatable modules. In a software defined radio prototype with two reconfigurable regions, the number of partial bitstreams is reduced by 50% and the compile time is shortened by 43%.In the present paper the background generation and motion detection algorithms, which are of key importance for the implementation of video detection, have been presented. A modification of the background generation algorithm, essential for proper algorithm functioning at medium and high road-traffic conditions, has been proposed. Algorithm adaptation for the implementation in reprogrammable device has been also presented. PixelStream-based implementation has been successfully performed. Real-time verification on reconfigurable platform has been done.

FPGA based controller for heterogenous image processing system

In the present paper construction of a controller is described, which is used to control the RETINA image processing platform. The 32-bit RETINA card is dedicated to be used for image acquisition, processing and analysis. The module resources include Video ADC, Virtex FPGA device, floating point Motorola 96002 DSP and PCI Master interface, what enables the execution of all the operations in real-time. Controller is the main control centre for the module, supervising the modes and phases of its operation, and it is also used as a arbiter for the modules communication resources.

Landmark detection for wearable home care ambient assisted living system

The importance of developing ambient assisted living technologies (AALT), intended to monitor elderly person activity in their home environment, increases in recent years. Among many approaches, systems based on the optical observation have became popular. Consequently, there is a need of developing robust and fast video analysis platforms and algorithms. In this paper, we propose the use of wearable cameras to monitor the daily activities of the elderly. The novelty of our approach is to analyze gathered data in real-time on wearable platform (without the need of storing video data) and consequently minimize privacy issues. We focus on the person localization in indoor environment, using natural landmark images taken whilst walking inside building during calibration procedure. The hypothesis that wearable video surveillance systems can be used to infer location (and thus detect and monitor the activity of an elderly person) has been confirmed by preliminary tests and results. However, still there are many issues to be solved and therefore further research and development of the proposed solution is planned.

Object detection and tracking for low-cost video surveillance system

Abstract Automated and intelligent video surveillance systems play important role in the modern world. Since the amount of various video streams that must be analyzed grows, such artificial intelligence systems can assist humans in performing tiresome tasks. As a result, the effectiveness of response to a dangerous situations is increasing (detect unexpected movement or unusual behavior that may pose a threat to people, property and infrastructure). Video surveillance systems have to meet several requirements: must be accurate and not produce too many false alarms, moreover it must be able to process the received video stream in real-time to provide a sufficient response time. The work presented here focuses on the selected challenges of scene analysis in video surveillance systems (object detection/tracking, effectiveness of the whole system). The aim of the research is to design a low-budget surveillance system, that can be used for example in a home security monitoring. Such solution can be use not only to surveillance but also to monitor elderly person at home or provide new ways of interacting in human-computer interaction systems.

Vision based facial action recognition system for people with disabilities

Human-computer interaction (HCI) is an emerging field of science aimed at providing natural ways for humans to use computers. There are various approaches to build mouse alternatives. Unfortunatelly, many existing solutions are not designed for people with severe physical disabilities. The work presented here is part of the larger project and focusses on facial action recognition framework and its components. It uses only optical observation - camera and image analysis and recognition algorithms. Proposed system gives possibility to serve the special needs of people with various disabilities, especially to use computers and possibly other devices. Consequently, in our opinion, it can offer disabled people access to a wide range of services which help them to participate more fully in society.

Eyetracking-based assessment of affect-related decay of human performance in visual tasks

Abstract Computerized recognition of human emotions enriches the man–machine interaction with a new personal and behavioral aspect. Firstly, because the influence of emotions on the human performance can be objectively assessed and predicted using standardized stimuli, secondly, because an emotion-sensitive computer system can be programmed to react adequately. This paper presents an investigation of emotional influence on the human scanpath. The proposed method assumes that the emotional state of the observer can distract his or her visual attention and can be reliably expressed in parameters of eye movements. We performed visual task experiments in order to record scanpaths from volunteers under auditory stress of controlled intensity. We used a calibrated set of stimuli and controlled the response of volunteers’ central nervous system expressed by the heart rate. For each efficiently stimulated participant, we related the scanpath parameters to the accuracy of solving a given task and to participant’s comment. As a principal result, we obtained a significant change of (1) saccade frequency in 90% and (2) maximum velocity in fixation phase in 80% of participants under stress. These results prove the influence of emotions on visual acuity, feasibility of eyetracking-based assessment of emotional state and motivate further investigations.

Low-Cost Scalable Home Video Surveillance System

Abstract Automated and intelligent video processing and analysis systems are becoming increasingly popular in video surveillance. Such systems must meet a number of requirements, such as threat detection and real-time video recording. Furthermore, they cannot be expensive and must not consume too much energy because they have to operate continuously. The work presented here focuses on building a home video surveillance system matching the household budget and possibly making use of hardware available in the house. Also, it must provide basic functionality (such as video recording and detecting threats) all the time, and allow for a more in-depth analysis when more computing power be available.

Pedestrian Detection and Analysis with Scale-Space and Distance Transform

Abstract Because the amount of various video streams recorded by video surveillance systems is increasing, the new approach, where human operator analyzing the video is replaced by artificial intelligence system is gaining new followers. The algorithm have to meet several requirements: must be accurate and not produce too many false alarms, moreover it must be able to process the received video stream in real-time to provide sufficient response time. In the article a system is presented which is able to detect and analyze walking pedestrians. It is based on two algorithms: scale space and matching contours using distance transform. The information can be used by other parts of the advanced video surveillance system, namely object tracking by detection, detecting heavy equipment only zone intrusion or for sorting out possible suspicious persons (pickpocket, homeless etc.).

Functional Endoscopic Sinus Surgery with Head Mounted Display and Video Analysis

Endoscopic sinus surgery (FESS) is the primary approach used today for the surgical treatment of many pathologies localized in the nasal cavity, paranasal sinuses, and, in select cases, in the base of the skull. Most surgeons currently prefer to operate using the video monitor rather than an eyepiece endoscope. Although video monitors have greatly enhanced surgical endoscopy by sharing the operation field view with nurses and anesthesiologists, video monitors require the operating surgeon to be focused on the screen instead of on the patient. In this article, the use of the Head-Mounted Display (HMD) is examined in functional endoscopic sinus surgery. The integration of imaging information from various sources (i.e. endoscope camera, CT scans) digitally is proposed using an image processing pipeline. The crucial requirement for the construction of a vision system is to ensure minimum video signal latency between the source and the display (HMD). Therefore, effects of various delays were examined during simulated surgical tasks. Also, an eye-tracking-based approach was used to quantitatively assess the time that the surgeon spent focusing on the screen versus observing the patient.

Ray-Tracing-Based Event Detection and 3D Visualization for Automated Video Surveillance System

Abstract Automated and intelligent video surveillance systems play important role in current home care and facilities security applications. Among many research problems is graphical visualization of semantic messages to the human operator that he can percept information in more natural way. The other essential research question is how to recognize 3D objects and their state on the monitored scene only from their views (2D images from the camera). In this paper we continue our previous work on data fusion in visualization of 3D scene semantic model and propose to recognize events and states of scene objects under surveillance in an automatic way using feedback provided by the renderer. We developed ray-tracing based visualization for surveillance system, that is capable of recognizing object’s state and at the same time present relevant information to the human operator.