Hrvoje Dujmić

Development of the communication/information infrastructure at the academic institution

The quality and functionality of the communication/information infrastructure is essential for the development of the academic institution. The status of the academic institution and its relationship with other subjects such as government institutions and companies establish the basic potentials and constraints for the infrastructural development. The Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture (FESB) of the University of Split in Croatia is an example of the academic institution with dynamic development of its communication/information infrastructure. The support by the Ministry of Science and Technology and the CARNet has been the major driving force in development of the basic communication/information infrastructure. The major result of this cooperation has been the deployment of the local area network and provision of the Internet connections and services. However, the intensive collaboration with the renown ICT companies on joint research and development projects in the area of ICT technology and services have generated new potentials for development of the communication/information infrastructure at FESB. Important results of this cooperation include installation of the digital PABX, development of the experimental communication network based on technologies such as ATM, ISDN, IP, xDSL, GSM and DECT. Another important step is the ongoing process of introduction of the wireless local area network to enable the mobility of students and faculty members within the premises. The plans for the future development of the communication/information infrastructure at FESB include continuous work on introduction of new ICT technologies and improvement of the functionality through the cooperation with the partners.

Local thresholding classified vector quantization with memory reduction

A new memory reduction method for classified vector quantization (CVQ) is presented. Symmetry reflection, rotation and inversion of edge subimages are used to join appropriate edge classes thus reducing the memory requirements for edge codebooks by 8(4) times for the classifier used in this paper. Besides the memory reduction, our method generates the more robust codebooks thus increasing the PSNR for images outside the training set. It also relieves codebook generation for high bit rate by reducing the number of images that should be inside the training set. The proposed method has been tested with a classifier that is based on the comparison of locally thresholded image vectors with a predefined set of binary edge templates.

Improved Linearized Combinatorial Model (ILCM) for optimal frame size selection in ALOHA-based RFID systems

Radio Frequency Identification (RFID) technology became the most important tool for identification of items and tracking. Nowadays, the most popular in terms of best price-performance ratio is passive RFID technology, where tags are both powered-up and communicating using the same radio waves transmitted via reader antenna(s).

Image segmentation based on complexity mining and mean-shift algorithm

Mean shift algorithm is a well established method for image segmentation. It is particularly popular technique due to non-parametric nature which enables efficient segmentation of complex arbitrary shapes. Despite such advantage, high computational complexity still makes it unsuitable for segmentation of high resolution images in time critical applications. This paper introduces a new approach which alleviates performance issues of mean shift using complexity reduction based on information theory. Proposed algorithm starts by calculating information potential field of the image in order to get insight into complexity of the regions. Afterwards, only complex regions are segmented by computationally expensive mean shift algorithm, while segmentation of simpler regions is performed by a cheaper method. Performance of our method is additionally improved with execution of the key code sections on the GPGPU platform. Experimental results have shown that our method produces comparable segmentation quality to regular parallel mean shift, but with significant reduction in overall execution time.

Multimedia dictionary network application: design and implementation

Proficiency in foreign languages is essential for professionals from countries in transition. A multimedia dictionary Internet application may offer needed help in everyday activities as well as in distance-learning systems. A multimedia dictionary is considered as both a conventional dictionary tool and a distance-learning network application. We analyze the architecture and implementation issues arising in the design and implementation of a multimedia dictionary network application. The application is considered multimedia since it is designed to support textual, visual, and audio information content. Practical implementation is shown to be feasible with current technology, but a lack of available multimedia standards has been recognized. This is why we have considered the existing standards as well as the standards in development. The application is developed in Java language, and may be run as either a standalone application or an applet. The latter entails accessibility in a ubiquitous manner to a large number of users. The application is intended for both non-native English users and professionals exposed to specific technical terminology.

Memory reduction and image quality enhancement method for classified vector quantization

This paper considers new memory reduction and image quality enhancement method for classified vector quantization (CVQ) using symmetry reflection, rotation and inversion of edge subimages. These are used to join appropriate edge classes thus reducing memory requirements for edge codebooks by 4(8) times. Besides the memory reduction and increases of PSNR for images outside the training set our method also relieves codebook generation for high bit rate by reducing the number of images that should be inside the training set. The proposed method has been tested with two different classification methods in order to ensure generality of the method.

Inverse error diffusion using table look-up and vector quantization: a novel scheme for low bit rate

This paper considers inverse error diffusion and data compression of binary error diffused images using table look-up and vector quantization (VQ). The encoding process needs a table look-up which transforms half-toned images to a set of codewords. The decoding process requires table look-up and low pass filtering. Using the proposed scheme the reconstructed gray-scale image is stored in a compressed form, Also, the error diffused image can be compressed in the gray-scale domain. Our method integrates the processes of inverse halftoning and compression and is useful when error diffusion halftoning algorithm is applied. The proposed scheme outperforms, for low bit rate (0.25 bpp to 0.5 bpp), similar existing techniques and is of low computational complexity.