We are investigating the use of badge based wearable computers to create highly mobile location and environment aware systems. When coupled to intelligent servers the badges provide an unparalleled platform for human centred information environments. This paper describes the architecture of the badge, its distributed computing environment, and presents initial results of application development trials conducted by a class of telecommunications students at KTH.
Lane detection is an important element of most driver assistance applications. A new lane detection technique that is able to withstand some of the common issues like illumination changes, surface irregularities, scattered shadows, and presence of neighboring vehicles is presented in this paper. At first, inverse perspective mapping and color space conversion is performed on the input image. Then, the images are cross-correlated with a collection of predefined templates to find candidate lane regions. These regions then undergo connected components analysis, morphological operations, and elliptical projections to approximate positions of the lane markers. The implementation of the Kalman filter enables tracking lane markers on curved roads while RANSAC helps improve estimates by eliminating outliers. Finally, a new method for calculating errors between the detected lane markers and ground truth is presented. The developed system showed good performance when tested with real-world driving videos containing variations in illumination, road surface, and traffic conditions.
Lane detection is an important application of driver assistance. In this paper, a new technique for detecting lane markers that is able to cope with many complex conditions is presented. Some of these conditions include dynamic illumination, scattered shadows, and the presence of neighboring vehicles to name a few. The input image is first pre-processed with a perspective removal transformation followed by a color space conversion. Then, the core elements of the proposed technique consisting of template matching, lane region merging, elliptical projections, and parametric tracking are explained. A formal error metric used in performance evaluation is also introduced. Finally, quantitative analyses show that the developed system performs well in real-world driving conditions with variations in illumination, traffic, and road surface quality.
In this paper, we present a new multi camera approach to Lane Departure Warning (LDW). Upon acquisition, the captured images are transformed to a bird's-eye view using a modified perspective removal transformation. Then, camera calibration is used to accurately determine the position of the two cameras relative to a reference point. Lane detection is performed on the front and rear camera images which are combined using data fusion. Finally, the distance between the vehicle and adjacent lane boundaries is determined allowing to perform LDW. The proposed system was tested on real world driving videos and shows good results when compared to ground truth.
In this paper, we present a new multi camera approach toLane DepartureWarning (LDW). First, a perspective removaltransformation is applied to the camera captured images toconvert them into bird’s-eye view images. Then, the positionof the two cameras relative to a reference point is accuratelydetermined using a new calibration technique. Lane detectionis performed on the front and rear camera images who resultsare combined using data fusion. Finally, LDW is implementedby determining the distance between the vehicle andadjacent lane boundaries. The proposed system was tested onreal world driving videos and shows good results when comparedto ground truth.
A new night-time lane detection system and its accompanying framework are presented in this paper. The accompanying framework consists of an automated ground truth process and systematic storage of captured videos that will be used for training and testing. The proposed Advanced Lane Detector 2.0 (ALD 2.0) is an improvement over the ALD 1.0 or Layered Approach with integration of pixel remapping, outlier removal, and prediction with tracking. Additionally, a novel procedure to generate the ground truth data for lane marker locations is also proposed. The procedure consists of an original process called time slicing, which provides the user with unique visualization of the captured video and enables quick generation of ground truth information. Finally, the setup and implementation of a database hosting lane detection videos and standardized data sets for testing are also described. The ALD 2.0 is evaluated by means of the user-created annotations accompanying the videos. Finally, the planned improvements and remaining work are addressed.
In this document, a new and efficient method to specify the ground truth locations of lane markers is presented. The method comprises of a novel process called Time-Slicing that provided the user with a unique visualization of the video. Coupled with automation via spline interpolation, the quick generation of necessary ground truth information is achieved. Videos recorded from a vehicle while driving on local city roads and highways are marked with ground truth information for use in testing. The performance of a variety of lane detection systems is compared to the ground truth and the error is computed for each system. Finally, quantitative analysis shows that the reference lane detection system presented in [1] produces the most accurate lane detections which is depicted by the smallest error.
In this paper, we present a new methodology for detecting lane markers that is able to withstand many challenging situations like scattered shadows, illumination changes, and presence of neighboring vehicles to name a few. At first, the input image undergoes a perspective removal followed by a color space conversion. Then, the core elements consisting of template matching, lane region merging, and elliptical projections are explained. Finally, the developed system showed good results when tested on 15 minutes of real-world driving videos containing variations in illumination, traffic, and road surface conditions.
A new night-time lane detection system that extends the idea of a Layered Approach [1] is presented in this document. The extension includes the incorporation of (1) inverse Perspective Mapping (IPM) to generate a bird's-eye view of the road surface, (2) application of Random Sample Consensus (RANSAC) to rid outliers from the data., and (3) Kalman filtering to smooth the output of the lane tracker. Videos of driving scenarios on local city roads and highways were used to test the new system. Quantitative analysis shows higher accuracy in detecting lane markers in comparison to other approaches.
In this paper, we propose a new methodology for detecting lane markers that exploits the parallel nature of lane boundaries on the road. First, the input image is pre-processed and filtered to detect lane marker features. Then, using the Polar Randomized Hough Transform that is introduced in this paper, lines are fitted through the detected features and the orientation of each line is evaluated. By finding near parallel lines separated by a constraint specified distance, false signalling caused by artifacts in the image is greatly reduced. The proposed system was tested using a real world driving videos and showed good results despite the presence of neighboring vehicles, shadows, and irregularities on the road surface.
In a previous paper, a simple approach to lane detection using the Hough transform and iterated matched filters was described [1]. This paper extends this work by incorporating an inverse perspective mapping to create a bird's-eye view of the road, applying random sample consensus to help eliminate outliers due to noise and artifacts in the road, and a Kalman filter to help smooth the output of the lane tracker.
A layered approach is designed to address many of the real-world problems that an inexpensive lane detection system would encounter. A region of interest is first extracted from the image followed by an enhancement procedure to manipulate the shape of the lane markers. The extracted region is then converted to binary using an adaptive threshold. A model based line detection system hypothesizes lane position. Finally, an iterated matched filtering scheme estimates the final lane position. The developed system shows good performance when tested on real-world data that contains fluctuating illumination and a variety of traffic conditions.
The mobility of portable computers and workstations is not transparent to users. They adjust to reduced services as long as they have no connection to a supporting infrastructure. The goal of the Walkstation project is to realize a user transparent mobile IP router (MINT) for wireless links (infrared and radio) operating at 1-10 Mbit/sec. For the study of user behavior and system characteristics, a campuswide testbed (ERIC) with 50-100 stations is planned to demonstrate the new solutions found in the Walkstation II project.
Named for the combination of the words DJ and Jamming, the DJammer allows its users to manipulate their music using standard DJ techniques as well as interact with others in virtual jam sessions through the exchange and sharing of multiple music streams. In this paper, we describe the evolution of the DJammer including the interactive process and evaluation which led to the latest prototype. This latest prototype allows its users to "touch" their digital music through single-handed manipulation and control. By allowing both creativity and communication with digital music, the DJammer is a new musical instrument which takes the next step in the evolution of portable music players.
Supporting a rich array of information sources is a key element to making highly mobile computing devices usable by the wider community. It is our belief that there will not be one specific killer application for this form of computing device, but an array of applications that the user can easily access. These applications will be context sensitive and associated with a range of activities. We have developed a custom watch platform that acts as a display for presenting this type of information to the user. In order to make the watch as small and low powered as possible, we have offloaded the processing onto an external mobile device we term a personal server, which is also carried by the user. We present our lightweight software infrastructure supporting a wrist-based display communicating with a portable personal server.
Ubiquitous mobile computing devices offer the opportunity to provide easy access to a rich set of information sources. Placing the display for this computing device on the user's wrist allows for quick, easy, and pervasive access to this information. In this paper we describe a user interface model and a set of five applications we have developed, with the aim of providing a user interface that supports lightweight interactions. Our goal is to make our pervasive watch as simple to use as a common wrist-watch worn today.
Third-generation mobile networks (3G) will support the end-to-end delivery of IP to wireless end-devices. The cost of this infrastructure and the moderate data-rates that can be expected have lead to proposals for a mixed 3G and wireless LAN (WLAN) approach. WLAN delivers much higher data-rates, currently up to 11 Mbps (IEEE 802.11b). Field trials with public WLAN extensions to Gigabit Ethernet networks show outdoor coverage of several hundred meters. Our previous work demonstrated smart delivery of multimedia involving agents running in the mobile, the access point, and the content provider. This allowed us to dynamically adapt both the application and network behavior (to each other) in order to meet the criteria for specific applications. In this paper, we extend this approach by adding service knowledge meta-data to the multimedia content (creating so-called Smart Media) to take advantage of the fact that for non real-time media content, which needs ample bandwidth to deliver, there does not need to be a coupling between transfer rate and playout rate. This approach enables the agent to further free resources for the delivery of streaming media to mobile users. In light of this approach, we propose novel network topologies with WLAN access using Smart Media Packets, for which we examine the minimal requirements for delivering the services.
Today, mobility of portable computers and workstations is not transparent to users. They adjust to reduced services as long as they have no connection to a supporting infrastructure. The goal of the Walkstation project is to realize a user transparent mobile IP router (MINT) for wireless links (infrared and radio) operating at 1-10 Mbit/sec. For the study of user behavior and system characteristics a campus wide testbed (ERIC) with 50-100 stations is planned to demonstrate the new solutions found in the Walkstation II project.
In the Swedish Walkstation project a device for mobile hosts and base stations is realized by the same kind of hardware but with different kinds of communication software. This device, called MINT (Mobile IP Router), consists mainly of three parts, one for the connection to the host or backbone net (Ethernet), one for connecting to wireless LANs (radio or infrared), and a processing part for computing the communication protocols. This device serves as more than a mobile modem, i.e., it actually routes packets over potentially multiple paths with varying connectivity and quality. By using Ethernet as an input channel, it should not be necessary to install special purpose wiring dedicated to mobile communication. Instead the existing network, which normally is globally available, is used to select suitable sites for base stations. To realize a large scale field trial the new router must be low cost, small in size, and have a low power consumption while offering high performance. The current phase of the project involves building a 68030 based router with multiple interfaces, including as host interfaces: SCSI, Ethernet, serial, parallel; and as wireless interfaces: IR, microwave-radio, or a-second Ethernet (supported for software development).
The goals of this project and a summary of what has been achieved thus far are presented. This includes a status report on our Mobile INTernet router which factors mobility out of the user device. Using this platform, we discuss the desirability of having multiple receivers, programmable radios, and new I/O devices. This paper will address how new technologies will enable future mobile distributed computing and communication systems. We seek to not only cut the umbilical cord which binds users to specific sites, but to allow users to easily make use of different infrastructures as necessary. This work is part of the Walkstation II project.
While new approaches to chemical localization have been proposed, animals are still widely used for locating landmines and illegal substances. Existing electronic noses still do not have the necessary sensitivity and accuracy. By modeling a cell's chemical detection system, we can gain insight into the basic "olfactory" system. We use an inspiration from chemotaxis and Hebbian learning to enhance localization and tracking of gradient sources, which can be applied to both chemicals and heat. The eukaryotic receptor clustering model shows improvement over previous prokaryotic chemotaxis-inspired methods that do not take into account receptor clustering. Receptor clustering essentially adapts receptors spatio-temporally. For a mobile simulation. our method locates the source in less convergence time than the other chemotaxis algorithms and insignificantly less time compared to no spatio-temporal filtering (e.g. a single-sensor memoryless case). We then show that local regions of receptor cooperation have the best performance reflecting observations of receptor behavior in biology. To demonstrate the performance of this system in real-time, a stationary 4/8-sensor version of the array is implemented, and the algorithm improves the convergence time, mean, and variance of the Direction-of-Arrival calculation in diffusive, turbulent, and noisy environments.
Gradient field localization, such as chemical and heat source location, is a complex problem, yet few designs have been proposed. Examples are locating fires, thermal leaks in insulation, explosive vapors, illegal substances, and chemical leaks. In this paper, we propose one 4 sensor stationary array to localize the direction-of-arrival (DOA) of a temperature source and model the approach on our previous biologically-inspired array processing technique. Several variants of the algorithm are shown, and with each degree of sophistication, it significantly reduces the DOA variance caused by noisy measurement conditions. Next, a clean 4/8-sensor version of the array was constructed, and it is shown that sensor cooperation improves the adaptation in diffusive, turbulent, and noisy environments.
The industrial design prototyping process has previously shown promising enhancements using Spatial Augmented Reality to increase the fidelity of concept visualizations. This paper explores further improvements to the process by incorporating tangible buttons to allow dynamically positioned controls to be employed by the designer. The tangible buttons are equipped with RFID tags that are read by a wearable glove sensor system to emulate button activation for simulating prototype design functionality. We present a new environmental setup to support the low cost development of an active user interface that is not restricted to the two-dimensional surface of a traditional computer display. The design of our system has been guided by the requirements of industrial designers and an expert review of the system was conducted to identify its usefulness and usability aspects. Additionally, the quantitative performance evaluation of the RFID tags indicated that the concept development using our system to support a simulated user interface functionality is an improvement to the design process.
Providing high resolution tangible user interface components without batteries such as dials and sliders that support dynamic user interface arrangement is challenging. Previous work uses RFID to support limited resolution custom-built components. We demonstrate improved techniques using commercial off the shelf input controls incorporated into passive RFID tags using an on-off key subcarrier to encode state information into the RFID signal. Our method supports high resolution components that do not require power cables or batteries. We provide exemplars demonstrating how the technique supports a range of user interface components including buttons, dials, sliders, flex and light sensors. Compared to previous work, we obtain a higher resolution, only limited by sample time, for all components and demonstrate 115 discrete dial positions. Our technique allows the TUI components to be freely placed and rearranged without hardwiring or batteries.
A number of different mobile payment solutions have been presented the last decade. The phone subscription with its security mechanisms are used for user identification and payments. This is the case for SMS based payment and ticketing systems that are getting more and more popular. However, there are other ways to implement a Trusted Element (TE) , where a SIM card architecture is only one. It can be in the mobile phone, as a separate integrated circuit, as an optional customer deployed plug-in device (e.g., microSD) or be running as an application on a server existing entirely as software.
In this paper we analyze what roles and responsibilities different actors have in different types of mobile payments solutions. The main focus is on the implications for the mobile operator business. It turns out that new types of intermediary actors in most cases play an important role. Sometimes mobile operators are not even involved. The emergence of new payment together with other non-SIM card based TE solutions opens up for many different market scenarios for mobile payment services.
This paper describes a product development collaborationbetween Swatch, a wearable design company, and HewlettPackard (HP), a provider of Information and CommunicationTechnologies (ICT). For both companies, the interaction wasunique in that it required them to consider the deployment ofwearable computing technology and support services into anapplication space dominated by fashion design. The author of thispaper worked at Hewlett Packard Laboratories, and wasextensively involved in this joint work. The goal of the projectwas to broaden the market for Swatch’s designs by using ICT.However as the project progressed it became clear that theinfluence of ICT on fashion design is highly complex andreciprocal. This paper describes what was observed during the 3years of the joint project, and presents ten learned outcomes fromthe attempt to merge wearable computing with fashion design.
This Innovate Practice Work-In-Progress paper describes a novel, knowledge sharing environment that addresses several current problems in technical education. This environment is called the Kista Mentorspace, and is hosted by the Swedish Royal Institute of Technology, KTH. Although the problems it deals with are many, this paper focuses on the following areas. 1. How to provision applied learning and practice to support technical theory. 2. How to do this in a way that shows convincing value from higher education institutions. 3. How to extend this value and contribution in a very broad sense. The Mentorspace provisions applied learning and practice by combining students from different schools in order to share their knowledge with engineering students. Knowledge sharing leads to applied practice including tool use, but that aspect is secondary. The Mentorspace increases the value of engineering education by tying applied practice to local needs and opportunities for graduates within the community. To extend this value in a broad sense, the Mentorspace brings in business and community developers to add local economic and social factors to the engineering curriculum. Success of the space is measured through academic and personal accomplishment, and contribution to community and social development. Long term mentoring by all participants regardless of age or social position is a cornerstone of the Mentorspace.
A system for identifying the position of a wireless device and transmitting that position back to the wireless device, the system comprising a plurality of receiver/correlators, each receiver/correlator being positioned at a fixed location in the network for receiving position request packets from the wireless cellular device, thereby generating a trigger signal each time a position request packet is received, the trigger signal used to record a local time, as indicated by an internal clock, at which the position request packet is received, and generating timing packets which include information about received position request packet, including time received. The system further including a central server for receiving the timing packets from the plurality of receiver/correlators and determining the position of the wireless device using the information in various timing packets.
A system for providing operational power to network access points includes a router having an external power supply to transmit network signals to the network access points. Each self-powered access point includes power extraction circuitry for extracting power from the network signals and a power storage subsystem for storing the extracted power. Signal conversion circuitry is connected to the power storage subsystem for receiving operating power to perform signal conversion of received network signals addressed to a computer and to reconfigure data received from the computer for transmission over the network. In a preferred embodiment, the power extraction circuitry includes a full-power extraction mode and a partial power extraction mode. A data analyzer monitors received network signals to determine whether the signals are required by the computer for processing. If a signal is required for processing, the data analyzer enables the partial power extraction mode in the power extraction circuitry to preserve a minimum signal power required by the computer for signal processing. If the signal is not required for processing, the data analyzer enables the full power extraction mode to extract substantially all power from the signal. Power level detection circuitry determines whether the power storage subsystem stores a minimum power level. If a power level below the minimum is detected, the power level detection circuitry disables the data analyzer and enables the full power extraction mode of the power extraction circuitry until the minimum power is achieved, at which time the data analyzer is re-enabled.
This is a case study that focuses on design methods developed by a computer engineer engaged in an interdisciplinary design project with a classic designer. Of interest is how the computer engineer initially found the task of achieving the shared design goals to be almost impossible, primarily due to the differences between the codified design methods of the engineer, and the tacit knowledge based methods of the designer. The study describes how the engineer developed new design realization skills enabling him to reconcile these differences in a way that allow the tacit knowledge of the designer to influence the codified engineering process in a repeatable way. These methods referred to as dreaming and mirroring represent a potentially learnable extension to the classic engineering design realization process.
Spatial Augmented Reality has shown promising results to support the industrial design process, this paper explores improvements by incorporating tangible buttons to allow dynamically positioned controls with a wearable glove sensor system for simulating prototype design functionality. We present a system to support the low cost development of an active user interface that is not restricted to the two-dimensional surface of a traditional computer display.