Catálogo de publicaciones - libros

A person identification system based on fast face learning of lifting wavelet filters is proposed. The real power of our system lies in fast learning of lifting wavelet filters adaptive to facial parts such as eyes, nose and lips, in a set of training faces. In our system, free parameters in the lifting filter are learned fast by using Newton’s method. The learned parameters are memorized in a database together with the training faces. The lifting filters with the learned parameters in the database are applied to each of video frames which contain faces of a person, and the faces are detected by measuring some kind of distance. A person whose face is detected in a maximum number of frames is identified as a target person. To realize fast face detection, the learned filters are applied only to the skin areas separated from background by using color segmentation. Simulation results show that our person identification algorithm is accurate and fast.

In this paper ideas of the dynamic character of design are presented by using intelligent agents to conceptual design aided by computer. These ideas are illustrated by the application related to decorative art. Two cooperative curious agents are used to generate periodic patterns. Agents’ curiosity controls evaluation of their actions, specification of new aims and planning future behavior. The interaction between agents and the environment strongly determines the course of designing.

The paper presents a pattern recognition study aimed on music notation recognition. The study is focused on practical aspect of optical music recognition; it presents a variety of methods applied in optical music recognition technology. The following logically separated stages of music notation recognition are distinguished: acquiring music notation structure, recognizing symbols of music notation, analyzing contextual information. The directions for OMR package development are drawn.

Most of the presented procedures were verified, the others are still under development. Representative set of 400 footprint measurements of tyres varying in construction, tread pattern, size, load, inflation pressure and degree of wear is used for verification.

Presented application gives practical benefit for tyre engineers who have to analyse many similar footprint images. It allows automatic assessment, classification and comparison of tyres on the basic of its footprint shape.

Several artificial intelligence methods of short-term electrical load forecasting are discussed in the paper. The model of a hybrid system based on syntactic pattern recognition, neural networks, and fuzzy techniques is introduced. The application of the model and the experimental results of short-term electrical load forecasting are presented.

This paper describes a fast and completely automatic algorithm for human face tracking. The tracked face is represented by a weighted histogram. The current histogram is compared to histograms at the particles’ positions. The weight of each particle is determined on the basis of Bhattacharyya distance and intensity gradient along the ellipse’s boundary. The incorporation of information about the distance between the camera and the face undergoing tracking results in robust tracking even in presence of skin colored regions in the background. The initialization of the tracker is realized by means of face detection. The detection is carried out using Haar-like features, followed by the verification of face distance to the camera and face region size heuristics.

Plucked sounds of musical instruments from chordophones group are examples of non-stationary sounds having both tonal and transient Plucked souncharacter. The experiments presented in this paper had Plucked sounto answer to the question if the wavelet packet transform based strategy of features extraction and selection that proved useful in many other classification tasks will be also useful for distinguishing differences of sounds produced by master quality violins played pizzicato.

Micro Electro Mechanical Systems will soon usher in a new technological renaissance. Just as ICs brought the pocket calculator, PC, and video games, MEMS will provide a new set of products and markets. Learn about the state of the art, from inertial sensors to microfluidic devices []. Over the last few years, considerable effort has gone into the study of the failure mechanisms and reliability of MEMS. Although still very incomplete, our knowledge of the reliability issues relevant to MEMS is growing. One of the major problems in MEMS production is fault detection. After fault diagnosis, hardware or software methods can be used to overcome it. Most of MEMS have nonlinear and complex models. So it is diffcult or impossible to detect the faults by traditional methods, which are model-based. In this paper different Neural Networks are used to classify and recognize faults. Different faults are recognized whilst considered as different patterns. We use different Neural Networks to classify different faults and fault free data. Two RF MEMS, which are RF Low pass filter and RF Inter digital capacitor are simulated by EM3DS, a MEMS software simulator. At last the results are compared.

In this paper we propose a hierarchical approach to representing recursive structures and their environment. Graph grammars are used to simulate the process of changes of each structure. This approach is combined with an artificial evolution that mimics the diversity of individuals within the same species. The proposed approach takes into consideration physical constraints of the real world. Our method is illustrated with examples of plant-like structures.

A new algorithm for recognition of symbols in license plates of Polish cars is presented. The method is aimed to work as the real-time system and successfully recognizes 60%–70% of plates. The algorithm utilizes image processing techniques and the neural network based approach. It was divided into three stages, the details of which are described in this paper. Presented algorithm shows that it is possible to construct an efficient videodetector, the abilities of which are comparable or better than the abilities of existing detectors, and which would cost not more than a personal computer.