Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
DISTORTED PATTERN RECOGNITION AND ANALYSIS WITH THE HELP OF Hy GRAPH REPRESENTATION
Full text linkAn algorithmfor distortedpattern recognition ispresented. It\u27s generalization ofM. Flasiń- ski results (Pattern Recognition, 27, 1-16, 1992). A new formalism allows to make both qualitative and quantitive distortion analysis. It also enlarges parser flexibility by exten- ding the set ofpatterns which may be recognized
SOFTWARE DEVELOPMENT COST ESTIMATION METHODS AND RESEARCH TRENDS
Full text linkEarly estirnation o f project size and completion time is essential for successful project plan- ning and tracking. Multiple methods have been proposed to estimate software size and cost parameters. Suitability of the estirnation methods depends on many factors like software application domain, product complexity, availability of historical data, team expertise etc. Most common and widely used estirnation technicjues are described and analyzed. Current research trends in software estirnation cost are also presented
FINITE ELEMENT COMPUTING USING NETWORK OF COMPUTERS AND PARALLEL MACHINES
No full textThis paper presents three parallel algorithms for simulation of axisymmetrical metal forging pro- cesses. The algorithms are based on the non-overlapping domain decomposition method where a mesh of elements is divided into layers assigned to different processes. Parallel programs imp- lementing these algorithms are written in C using PVM calls. The efficiency of the parallel pro grams is verified on networked IBM RS/6000, Sun SPARCstations and HP 9000 workstations as well as on HP/Convex Exemplar SPP1000
The Internet Inter-Orb Protocol Security Bridge
No full textThe development of the CORBA standard and the creation of the IIOP protocol allowed the creation of distributed object-oriented applications in heterogeneous systems. The IIOP protocol enabled communication between brokers\u27 requests from various manufacturers. It also began to be used as a protocol for communication of Java applet objects run in a web browser environment with objects running on the server, replacing the less flexible and universal CGI (Common Gateway Interface) protocol. Therefore, it became deliberate to provide CORBA facilities on the Internet. However, sharing these facilities requires two solutions. The first problem concerns networks protected by a firewall. Some of these devices allow communication through one or more TCP and UDP ports, which makes it impossible to use the IIOP protocol directly. The solution to this problem is tunneling of the IIOP (HTTP tunneling) protocol through other protocols; however, this is an unusual and not recommended approach. Another way to make the object available is to start the object so that its object adapter is running on a specific TCP port open through the firewall. You can also run a proxy (proxy) of the shared object on the port open through the firewall. The second problem concerns the provision of object access control. In the case of the IIOP protocol, firewall protection is not sufficient when it is done at a level lower than the IIOP protocol layer. Control of access to CORBA objects should be implemented using the concepts of CORBA and IIOP standards. This requires the creation of a system operating at the level of these layers, cooperating with the fire wall
KNOWLEDGE REPRESENTATION AND INFERENCE FOR ANALYSIS AND DESIGN OF DATABASES AND TABULAR RULE-BASED SYSTEMS
Full text linkRulc-based Systems constitute a powerful tool for speciftcation of knowledge in design and implementation of knowledge-based Systems. They provide also a universal programming paradigm for domains such as intelligent control, decision support, situation classification and opcrational knowledge encoding. In order to assure safe and reliable performance, such Systems should satisfy certain format reąuirements, including completeness and consistency. This paper addresses the issue of analysis and verification of selected properties of a class of such Systems in a systematic way. A uniform, tabular scheme of single-levcl rule-bascd Systems is considered. Such systcms can be applied as a generalized form of databases for speciftcation of data pattems (unconditional knowledge), or can be used for deftning attributive decision tables (conditional knowledge in form of rules). They can also serve as lower-level componcnts of a hierarchical, multi-lcvcl control and decision support knowledge-based systcms. An algebraic knowledge rcprescntation paradigm using extcnded tabular rcprcsentation, similar to relational databasc tables is prcsentcd and algebraic bascs for system analysis, vcrification and design support arc outlined
LEXICAL SERVER OF POLISH LANGUAGE
No full textThis paper prcsents Lcxical Server of Polish Language, the tool that aids natural language Processing (NLP). Databasc of the scrver consists of dictionary units enrichcd by flcxical information. The lexical scrver should be able to perform identification of word forms and generations of all inflectcd forms of the word. The server is dedicated to the peoplc who are looking for NLP algorithms or implcmcnt them. The algorithms can be implemcntcd in different kinds of programming languages and different operating systems. Thcrc are somc cxamples of problems when lexical scrver can be useful: automatic text correction, tcxt indexing, keywords cxtraction, text profile building
ARTIFICIAL INTELLIGENCE METHODS IN ANALYSIS OF MORPHOLOGY OF SELECTED STRUCTURES IN MEDICAL IMAGES
No full textThe goal of this paper is the presentation of the possibilitics of application of syntactic methods of Computer image analysis for recognition of local stenoses of coronary arteries lumen and dctcction of pathological signs in upper parts of ureter ducts and rcnal calyxcs. Analysis of correct morphology of these structures is possiblc thanks to the application of sequcncc and trec methods from the group of syntactic methods of pattem recognition. In the case of analysis of coronary arteries images the main objective is computer-aidcd carly diagnosis of different forms of ischcmic cardiovascular diseases. Such diseases may rcvcal in the form of stablc or unstablc disturbanccs of heart rhythm or infaretion. In analysis of kidney radiograms the main goal is recognition of local irrcgularities in ureter lumens and examination of morphology of rcnal pc!vis and calyxcs
FAST AUTOMATIC CONFIGURATION OF ARTIFICIAL NEURAL NETWORKS USED FOR BINARY PATTERNS RECOGNITION
Full text linkThis paper prcsents a powcrful method of an automatically generated architccturc of ncural networks uscd for binary pattems rccognition, which can quickly and automatically rcducc synapses in a way of minimally reducing a quality of rccognition and a quality of generalization. Moreovcr, this method computes all weights in two runs over a leaming sequence, what makes this method vcry fast. First, the method calculatcs all binary featureś for each pattem and then weights are computcd. Furthcrmore, there is a quality of generalization considered because it is one o f the most important factors o f recognition whilc using neural networks
AUTOMATIC CLASSIFICATION OF NOUNS INTO SEMANTIC GROUPS USING A CORPUS OF TEXTS
Full text linkThis article presents a method of classification of nouns into semantic groups based on statistical inference. The algorithm uses the infiectional dictionary o f the Polish language and a corpus of texts to ana yse adjective-noun relationships. The semantic groups are consistent with the categorization in the WordNet dictionary. The classification of nouns into semantic groups is a smali step towards constructing a semantic dictionary for the Polish language
Salt and pepper noise reduction and edge detection algorithm based on neutrosophic logic
Full text linkNeutrosophic set (NS) is a powerful tool to deal with indeterminacy. In this paper, the neutrosophic set is applied to the image domain and a novel edge detection technique is proposed. Noise reduction of images is a challenging task in image processing. Salt and pepper noise is one kind of noise that affects a grayscale image significantly. Generally, the median filter is used to reduce salt and pepper noise; it gives optimum results while compared to other image filters. Median filter works only up to a certain level of noise intensity. Here we proposed a neighborhood-based image filter called nbd-filter, it works perfectly for gray image regardless of noise intensity. It reduces salt and pepper noise significantly at any noise level and produces a noise-free image. Further, we proposed an edge detection algorithm based on the neutrosophic set, it detectsedges efficiently for images corrupted by noise and noise-free images. Since most of the real-life images consists of indeterminate regions, neutrosophy is a perfect tool for edge detection. The main advantage of the proposed edge detector is, it is a simple and efficient technique and detect edges more efficient than conventional edge detectors