Real-Time Recognition System for Traffic Signs

The aim of this thesis project is to develop the Traffic Sign Recognition algorithm for real time. Inreal time environment, vehicles move at high speed on roads. For the vehicle intelligent system itbecomes essential to detect, process and recognize the traffic sign which is coming in front ofvehicle with high relative velocity, at the right time, so that the driver would be able to pro-actsimultaneously on instructions given in the Traffic Sign. The system assists drivers about trafficsigns they did not recognize before passing them. With the Traffic Sign Recognition system, thevehicle becomes aware of the traffic environment and reacts according to the situation.The objective of the project is to develop a system which can recognize the traffic signs in real time.The three target parameters are the system’s response time in real-time video streaming, the trafficsign recognition speed in still images and the recognition accuracy. The system consists of threeprocesses; the traffic sign detection, the traffic sign recognition and the traffic sign tracking. Thedetection process uses physical properties of traffic signs based on a priori knowledge to detect roadsigns. It generates the road sign image as the input to the recognition process. The recognitionprocess is implemented using the Pattern Matching algorithm. The system was first tested onstationary images where it showed on average 97% accuracy with the average processing time of0.15 seconds for traffic sign recognition. This procedure was then applied to the real time videostreaming. Finally the tracking of traffic signs was developed using Blob tracking which showed theaverage recognition accuracy to 95% in real time and improved the system’s average response timeto 0.04 seconds. This project has been implemented in C-language using the Open Computer VisionLibrary.

Hand Gesture Detection & Recognition System

The project introduces an application using computer vision for Hand gesture recognition. A camera records a live video stream, from which a snapshot is taken with the help of interface. The system is trained for each type of count hand gestures (one, two, three, four, and five) at least once. After that a test gesture is given to it and the system tries to recognize it.A research was carried out on a number of algorithms that could best differentiate a hand gesture. It was found that the diagonal sum algorithm gave the highest accuracy rate. In the preprocessing phase, a self-developed algorithm removes the background of each training gesture. After that the image is converted into a binary image and the sums of all diagonal elements of the picture are taken. This sum helps us in differentiating and classifying different hand gestures.Previous systems have used data gloves or markers for input in the system. I have no such constraints for using the system. The user can give hand gestures in view of the camera naturally. A completely robust hand gesture recognition system is still under heavy research and development; the implemented system serves as an extendible foundation for future work.

An Upgrade of Network Traffic Recognition System for SIP/VoIP Traffic Recognition

The purpose of this project is to update the tool of Network Traffic Recognition System (NTRS) which is proprietary software of Ericsson AB and Tsinghua University, and to implement the updated tool to finish SIP/VoIP traffic recognition. Basing on the original NTRS, I analyze the traffic recognition principal of NTRS, and redesign the structure and module of the tool according to characteristics of SIP/VoIP traffic, and then finally I program to achieve the upgrade. After the final test with our SIP data trace files in the updated system, a satisfactory result is derived. The result presents that our updated system holds a rate of recognition on a confident level in the SIP session recognition as well as the VoIP call recognition. In the comparison with the software of Wireshark, our updated system has a result which is extremely close to Wireshark’s output, and the working time is much less than Wireshark. In the aspect of practicability, the memory overflow problem is avoided, and the updated system can output the specific information of SIP/VoIP traffic recognition, such as SIP type, SIP state, VoIP state, etc. The upgrade fulfills the demand of this project.

Demonstration of Solar Heating and Cooling System using Sorption Integrated Solar Thermal Collectors

Producing cost-competitive small and medium-sized solar cooling systems is currently a significant challenge. Due to system complexity, extensive engineering, design and equipment costs; the installation costs of solar thermal cooling systems are prohibitively high. In efforts to overcome these limitations, a novel sorption heat pump module has been developed and directly integrated into a solar thermal collector. The module comprises a fully encapsulated sorption tube containing hygroscopic salt sorbent and water as a refrigerant, sealed under vacuum with no moving parts. A 5.6m2 aperture area outdoor laboratory-scale system of sorption module integrated solar collectors was installed in Stockholm, Sweden and evaluated under constant re-cooling and chilled fluid return temperatures in order to assess collector performance. Measured average solar cooling COP was 0.19 with average cooling powers between 120 and 200 Wm-2 collector aperture area. It was observed that average collector cooling power is constant at daily insolation levels above 3.6 kWhm-2 with the cooling energy produced being proportional to solar insolation. For full evaluation of an integrated sorption collector solar heating and cooling system, under the umbrella of a European Union project for technological innovation, a 180 m2 large-scale demonstration system has been installed in Karlstad, Sweden. Results from the installation commissioned in summer 2014 with non-optimised control strategies showed average electrical COP of 10.6 and average cooling powers between 140 and 250 Wm-2 collector aperture area. Optimisation of control strategies, heat transfer fluid flows through the collectors and electrical COP will be carried out in autumn 2014.

Road and Traffic Signs Recognition using Vector Machines

Intelligent Transportation System (ITS) is a system that builds a safe, effective and integrated transportation environment based on advanced technologies. Road signs detection and recognition is an important part of ITS, which offer ways to collect the real time traffic data for processing at a central facility.This project is to implement a road sign recognition model based on AI and image analysis technologies, which applies a machine learning method, Support Vector Machines, to recognize road signs. We focus on recognizing seven categories of road sign shapes and five categories of speed limit signs. Two kinds of features, binary image and Zernike moments, are used for representing the data to the SVM for training and test. We compared and analyzed the performances of SVM recognition model using different features and different kernels. Moreover, the performances using different recognition models, SVM and Fuzzy ARTMAP, are observed.

Road Sign Recognition based onInvariant Features using SupportVector Machine

Since last two decades researches have been working on developing systems that can assistsdrivers in the best way possible and make driving safe. Computer vision has played a crucialpart in design of these systems. With the introduction of vision techniques variousautonomous and robust real-time traffic automation systems have been designed such asTraffic monitoring, Traffic related parameter estimation and intelligent vehicles. Among theseautomatic detection and recognition of road signs has became an interesting research topic.The system can assist drivers about signs they don’t recognize before passing them.Aim of this research project is to present an Intelligent Road Sign Recognition System basedon state-of-the-art technique, the Support Vector Machine. The project is an extension to thework done at ITS research Platform at Dalarna University [25]. Focus of this research work ison the recognition of road signs under analysis. When classifying an image its location, sizeand orientation in the image plane are its irrelevant features and one way to get rid of thisambiguity is to extract those features which are invariant under the above mentionedtransformation. These invariant features are then used in Support Vector Machine forclassification. Support Vector Machine is a supervised learning machine that solves problemin higher dimension with the help of Kernel functions and is best know for classificationproblems.

Assessment of PD Speech Anomalies @ Home

Background: Voice processing in real-time is challenging. A drawback of previous work for Hypokinetic Dysarthria (HKD) recognition is the requirement of controlled settings in a laboratory environment. A personal digital assistant (PDA) has been developed for home assessment of PD patients. The PDA offers sound processing capabilities, which allow for developing a module for recognition and quantification HKD. Objective: To compose an algorithm for assessment of PD speech severity in the home environment based on a review synthesis. Methods: A two-tier review methodology is utilized. The first tier focuses on real-time problems in speech detection. In the second tier, acoustics features that are robust to medication changes in Levodopa-responsive patients are investigated for HKD recognition. Keywords such as Hypokinetic Dysarthria , and Speech recognition in real time were used in the search engines. IEEE explorer produced the most useful search hits as compared to Google Scholar, ELIN, EBRARY, PubMed and LIBRIS. Results: Vowel and consonant formants are the most relevant acoustic parameters to reflect PD medication changes. Since relevant speech segments (consonants and vowels) contains minority of speech energy, intelligibility can be improved by amplifying the voice signal using amplitude compression. Pause detection and peak to average power rate calculations for voice segmentation produce rich voice features in real time. Enhancements in voice segmentation can be done by inducing Zero-Crossing rate (ZCR). Consonants have high ZCR whereas vowels have low ZCR. Wavelet transform is found promising for voice analysis since it quantizes non-stationary voice signals over time-series using scale and translation parameters. In this way voice intelligibility in the waveforms can be analyzed in each time frame. Conclusions: This review evaluated HKD recognition algorithms to develop a tool for PD speech home-assessment using modern mobile technology. An algorithm that tackles realtime constraints in HKD recognition based on the review synthesis is proposed. We suggest that speech features may be further processed using wavelet transforms and used with a neural network for detection and quantification of speech anomalies related to PD. Based on this model, patients' speech can be automatically categorized according to UPDRS speech ratings.

Dialekter och röstigenkänning : Ett röstigenkännings-API:s förmåga att uppfatta svenska dialekters kännetecken och röstkombinationer

Allt eftersom utvecklingen går framåt inom applikationer och system så förändras också sättet på vilket vi interagerar med systemet på. Hittills har navigering och användning av applikationer och system mestadels skett med händerna och då genom mus och tangentbord. På senare tid så har navigering via touch-skärmar och rösten blivit allt mer vanligt. Då man ska styra en applikation med hjälp av rösten är det viktigt att vem som helst kan styra applikationen, oavsett vilken dialekt man har. För att kunna se hur korrekt ett röstigenkännings-API (Application Programming Interface) uppfattar svenska dialekter så initierades denna studie med dokumentstudier om dialekters kännetecken och ljudkombinationer. Dessa kännetecken och ljudkombinationer låg till grund för de ord vi valt ut till att testa API:et med. Varje dialekt fick alltså ett ord uppbyggt för att vara extra svårt för API:et att uppfatta när det uttalades av just den aktuella dialekten. Därefter utvecklades en prototyp, närmare bestämt en android-applikation som fungerade som ett verktyg i datainsamlingen. Då arbetet innehåller en prototyp och en undersökning så valdes Design and Creation Research som forskningsstrategi med datainsamlingsmetoderna dokumentstudier och observationer för att få önskat resultat. Data samlades in via observationer med prototypen som hjälpmedel och med hjälp av dokumentstudier. Det empiriska data som registrerats via observationerna och med hjälp av applikationen påvisade att vissa dialekter var lättare för API:et att uppfatta korrekt. I vissa fall var resultaten väntade då vissa ord uppbyggda av ljudkombinationer i enlighet med teorin skulle uttalas väldigt speciellt av en viss dialekt. Ibland blev det väldigt låga resultat på just dessa ord men i andra fall förvånansvärt höga. Slutsatsen vi drog av detta var att de ord vi valt ut med en baktanke om att de skulle få låga resultat för den speciella dialekten endast visade sig stämma vid två tillfällen. Det var istället det ord innehållande sje- och tje-ljud som enligt teorin var gemensamma kännetecken för alla dialekter som fick lägst resultat överlag.

Fingerprint Segmentation

In this thesis, a new algorithm has been proposed to segment the foreground of the fingerprint from the image under consideration. The algorithm uses three features, mean, variance and coherence. Based on these features, a rule system is built to help the algorithm to efficiently segment the image. In addition, the proposed algorithm combine split and merge with modified Otsu. Both enhancements techniques such as Gaussian filter and histogram equalization are applied to enhance and improve the quality of the image. Finally, a post processing technique is implemented to counter the undesirable effect in the segmented image. Fingerprint recognition system is one of the oldest recognition systems in biometrics techniques. Everyone have a unique and unchangeable fingerprint. Based on this uniqueness and distinctness, fingerprint identification has been used in many applications for a long period. A fingerprint image is a pattern which consists of two regions, foreground and background. The foreground contains all important information needed in the automatic fingerprint recognition systems. However, the background is a noisy region that contributes to the extraction of false minutiae in the system. To avoid the extraction of false minutiae, there are many steps which should be followed such as preprocessing and enhancement. One of these steps is the transformation of the fingerprint image from gray-scale image to black and white image. This transformation is called segmentation or binarization. The aim for fingerprint segmentation is to separate the foreground from the background. Due to the nature of fingerprint image, the segmentation becomes an important and challenging task. The proposed algorithm is applied on FVC2000 database. Manual examinations from human experts show that the proposed algorithm provides an efficient segmentation results. These improved results are demonstrating in diverse experiments.

Developing a Communication link between Agents and cross Platform IDE

The main objective of this thesis work is to develop communication link between Runrev Revolution (IDE) and JADE (Multi-Agent System) through Socket programming using TCP/IP layer. These two independent platforms are connected using socket programming technique. Socket programming is considered to be newly emerging technology among these two platforms, the work done in this thesis work is considered to be a prototype.A Graphical simulation model is developed by salixphere (Company in Hedemora) to simulate logistic problems using Runrev Revolution (IDE). The simulation software/program is called “BIOSIM”. The logistic problems are complex, and conventional optimization techniques are unlikely very successful. “BIOSIM” can demonstrate the graphical representation of logistic problems depending upon the problem domains. As this simulation model is developed in revolution programming language (Transcript) which is dynamically typed and English-like language, it is quite slow compared to other high level programming languages. The object of this thesis work is to add intelligent behaviour in graphical objects and develop communication link between Runrev revolution (IDE) and JADE (Multi-Agent System) using TCP/IP layers.The test shows the intelligent behaviour in the graphical objects and successful communication between Runrev Revolution (IDE) and JADE (Multi-Agent System).

Use of Photovoltaic on an E-bike? : A Feasibility Study

In recent years the number of bicycles with e-motors has been increased steadily. Within the pedelec – bikes where an e-motor supports the pedaling – a special group of transportation bikes has developed. These bikes have storage boxes in addition to the basic parts of a bike. Due to the space available on top of those boxes it is possible to install a PV system to generate electricity which could be used to recharge the battery of the pedelec. Such a system would lead to grid independent charging of the battery and to the possibility of an increased range of motor support. The feasibility of such a PV system is investigated for a three wheeled pedelec delivered by the company BABBOE NORDIC.The measured data of the electricity generation of this mobile system is compared to the possible electricity generation of a stationary system.To measure the consumption of the pedelec different tracks are covered, and the energy which is necessary to recharge the bike battery is measured using an energy logger. This recharge energy is used as an indirect measure of the electricity consumption. A PV prototype system is installed on the bike. It is a simple PV stand alone system consisting of PV panel, charge controller with MPP tracker and a solar battery. This system has the task to generate as much electricity as possible. The produced PV current and voltage aremeasured and documented using a data logger. Afterwards the average PV power is calculated. To compare the produced electricity of the on-bike system to that of a stationary system, the irradiance on the latter is measured simultaneously. Due to partial shadings on the on-bike PV panel, which are caused by the driver and some other bike parts, the average power output during riding the bike is very low. It is too low to support the motor directly. In case of a similar installation as the PV prototype system and the intention always to park the bike on a sunny spot an on-bike system could generate electricity to at least partly recharge a bike battery during one day. The stationary PV system using the same PV panel could have produced between 1.25 and 8.1 times as much as the on-bike PV system. Even though the investigation is done for a very specific case it can be concluded that anon-bike PV system, using similar components as in the investigation, is not feasible to recharge the battery of a pedelec in an appropriate manner. The biggest barrier is that partial shadings on the PV panel, which can be hardly avoided during operation and parking, result in a significant reduction of generated electricity. Also the installation of the on-bike PV system would lead to increased weight of the whole bike and the need for space which is reducing the storage capacity. To use solar energy for recharging a bike battery an indirect way is giving better results. In this case a stationary PV stand alone system is used which is located in a sunny spot without shadings and adjusted to use the maximum available solar energy. The battery of the bike is charged using the corresponding charger and an inverter which provides AC power using the captured solar energy.

Emissions from realistic operation of residential wood pellets heating systems

Emissions from residential combustion appliances vary significantly depending on the firing behaviours and combustion conditions, in addition to combustion technologies and fuel quality. Although wood pellet combustion in residential heating boilers is efficient, the combustion conditions during start-up and stop phases are not optimal and produce significantly high emissions such as carbon monoxide and hydrocarbon from incomplete combustion. The emissions from the start-up and stop phases of the pellet boilers are not fully taken into account in test methods for ecolabels which primarily focus on emissions during operation on full load and part load. The objective of the thesis is to investigate the emission characteristics during realistic operation of residential wood pellet boilers in order to identify when the major part of the annual emissions occur. Emissions from four residential wood pellet boilers were measured and characterized for three operating phases (start-up, steady and stop). Emissions from realistic operation of combined solar and wood pellet heating systems was continuously measured to investigate the influence of start-up and stop phases on total annual emissions. Measured emission data from the pellet devices were used to build an emission model to predict the annual emission factors from the dynamic operation of the heating system using the simulation software TRNSYS. Start-up emissions are found to vary with ignition type, supply of air and fuel, and time to complete the phase. Stop emissions are influenced by fan operation characteristics and the cleaning routine. Start-up and stop phases under realistic operation conditions contribute 80 – 95% of annual carbon monoxide (CO) emission, 60 – 90% total hydrocarbon (TOC), 10 – 20% of nitrogen oxides (NO), and 30 – 40% particles emissions. Annual emission factors from realistic operation of tested residential heating system with a top fed wood pelt boiler can be between 190 and 400 mg/MJ for the CO emissions, between 60 and 95 mg/MJ for the NO, between 6 and 25 mg/MJ for the TOC, between 30 and 116 mg/MJ for the particulate matter and between 2x10-13 /MJ and 4x10-13 /MJ for the number of particles. If the boiler has the cleaning sequence with compressed air such as in boiler B2, annual CO emission factor can be up to 550 mg/MJ. Average CO, TOC and particles emissions under realistic annual condition were greater than the limits values of two eco labels. These results highlight the importance of start-up and stop phases in annual emission factors (especially CO and TOC). Since a large or dominating part of the annual emissions in real operation arise from the start-up and stop sequences, test methods required by the ecolabels should take these emissions into account. In this way it will encourage the boiler manufacturers to minimize annual emissions. The annual emissions of residential pellet heating system can be reduced by optimizing the number of start-ups of the pellet boiler. It is possible to reduce up to 85% of the number of start-ups by optimizing the system design and its controller such as switching of the boiler pump after it stops, using two temperature sensors for boiler ON/OFF control, optimizing of the positions of the connections to the storage tank, increasing the mixing valve temperature in the boiler circuit and decreasing the pump flow rate. For 85 % reduction of start-ups, 75 % of CO and TOC emission factors were reduced while 13% increase in NO and 15 % increase in particle emissions was observed.