This study presents a complete octree construction algorithm for 2D depth images obtained from orthogonal face views, which can represent 3D objects exactly. In constructing quadtree, optimal quadtree construction algorithm is applied to depth images for efficient use of memory and reduction of tree construction time. In addition, pseudo-octrees are constructed by using our proposed method, which construct pseudo-octrees according to the resolution value given in each node of constructed quadtree and mapping relation between quadrants and octants. Finally, a complete octree, which represents a 3D object, is constructed by volume intersection with each constructed pseudo-octree. The representation accuracy of a complete octree constructed by our algorithm is investigated by using a 3D display method and a volume ratio method for a complete octree.

This paper is to propose an analytical queuing model for parallel processing of sequential program with nested loops. The analytical results are compared with the results from the implemented multiprocessor system composed of four intel 8088 microprocessor, eight 2KB shared common memories, and a hardware token ring. At results, this study shows that the processed results are almost similar in proposed analytical model and real system. Proposed analytical model can be applied to evaluate parallel processing of sequential program with nested loops.

In this paper, a rule checker which aids in composing a consistent knowledge base by checking redundancies and inconsistencies in a knowledge base is proposed. The proposed algorithm checks the rules by representing the rule connections as a network . The standard model of the rules adapted in this algorithm is in the Conjunctive Normal Form which includes NOT's, and rules of conventional expert system can be checked by converting them into the standard form by a rule form at converter. When compared with Ginsberg's KB-reducer which is conceptually most similar to the proposed algorithm among existing methods,it is shown by a computer simulation that with 360 rules, the checking time is three times faster and the rate increased as the number of rules increased, but the total memory requirement of the proposed agorithm is 1.2 times larger. The proposed algorithm has further advantages in that it can check circular rule chains and can find the paths of the redundant and inconsistent rules.

This paper presents lattice structure of bilinear filter and the conversion equations from lattice parameters to direct-form parameters. Billnear models are attractive for adaptive filtering applications because they can approximate a large class of nonlinear systems adequately, and usually with considerable parsimony in the number of coefficients required. The lattice filter formulation transforms the nonlinear filtering problem into an equivalent multichannel linear filtering problem and then uses multichannel lattice filtering algorithms to solve the nonlinear filtering problem. The lattice filters perform a Gram-Schmidt orthogonalization of the input data and have very good easily extended to more general nonlinear output feedback structures.

This paper presents two fast least-squares lattice algorithms for adaptive nonlinear filters equipped with bilinear system models. The lattice filters perform a Gram-Schmidt orthogonalization of the input data and have very good numerical properties. Furthermore, the computational complexity of the algorithms is an order of magnitude snaller than previously algorithm is an order of magnitude smaller than previously available methods. The first of the two approaches is an equation error algorithm that uses the measured desired response signal directly to comprte the adaptive filter outputs. This method is conceptually very simple`however, it will result in biased system models in the presence of measurement noise. The second approach is an approximate least-squares output error solution. In this case, the past samples of the output of the adaptive system itself are used to produce the filter output at the current time. Results of several experiments that demonstrate and compare the properties of the adaptive bilinear filters are also presented in this paper. These results indicate that the output error algorithm is less sensitive to output measurement noise than the squation error method.

As a result of recent progress in speech synthesis techniques, the those new services using new techniques are going to introduce into the telephone communication system. In setting standards, voice quality is obviously an important criterion. It is very important to develope a quality evaluation method of synthesized speech for the diagnostic assessment of system algorithm, and fair comparison of assessment values. This paper has described several basic concepts and criterions for quality assessment (intelligibility) of synthesized speech by rule, and then a word selection method and the word list to be used in word intelligibility test were proposed. Finally, a test method for word intelligibility is described.

This paper proposes and analyzes a new CFAR(Constant False Alarm Rate) detector called the EXGO(Excision Greatest Of)-CFAR. This is the combination of the EXCA(Excision Cell Averaging)-CFAR that shows a good performance under the influence of interferences and the GO(Greatest Of)-CFAR that fights well with clutter edges. For the performance analysis, the formulas for the detection probability and the false alarm probability are derived and computed, and the results are compared with other existing CFAR detectors. Our analysis shows that the proposed EXGO-CFAR considerably improves the false-alarm-rate performance of the EXCA-CFAR at clutter edges while maintaining the high detection probability performance of the EXCA-CFAR in the homogeneous and/or interference noise environment.

This paper proposes a neuro-controller for position control of one-link flexible robot arm. Basically the controller consists of a multi-layer neural network and a conventional PD controller. Two controller are parallelly connected. Neural network is traind by the conventional error back propagation learning rules. During learning period, the weights of neural network are adjusted to minimize the position error between the desired hub angle and the actual one. Finally the effectiveness of the proposed approach will be demonstrated by computer simulation.

A new method of driving the PDP(Plasma Display Panel) was proposed and its characteristics were investigated. Applying high frequency non-discharge pulses to an auxiliary anode resulted an increased region of stable operation, decreased delay time and increased light intensity. It is suggested that PDP with DPM (Double Pulse Memory) drive technique improves the delay time, luminance, region of stable operation and luminous efficacy compared to PDP with PPM(Planar Pulse Memory) drive technique developed by NHK Lab. in Japan.

The vision system is essential for SMT(Surface Mounting Technology) automation. The system plays the role of matching the positions betweem SMD and PCB, and inspecting SMD in the final stage of mounting. Real-time processing and high-precision are indispensable for practical purpose. In this paper, a new algorithm for position matching and inspection of SMD is proposed, and which is implemented on the DSP board using DSP board using DSP5600. Experimental results show mean matching error within 0.1 mm in the direction of x,y and execution time within 300msec. Therefore, we could attain high-speed and high-precision of the vision system for SMT automation.

In this paper, we propose a parameter-free smoothing method for speckle images, i.e., an adaptive least squares image smoothing technique implemented in a multistep environment. The pertinent smoothing window size at a given pixel is determined by the discontinuity measure which is defined by the ratio of the local variance and mean squares of intensity values of pixels over the smoothing window centered there. The mode of the discontinuity measure at each step is estimated to replace the noise variance parameter that is required in the adaptive smoothing. Computer simulation shows that the proposed multistep technique can smooth homogeneous regions satisfactorily while preserving fine details near boundaries.

The multilayer perceptron, trained by the error back-propagation learning rule, has been known as a mapping network which can represent arbitrary functions. However depending on the complexity of a function and the initial weights of the multilayer perceptron, the error back-propagation learning may fall into a local minimum or a flat area which may require a long learning time or lead to unsuccessful learning. To solve such difficulties in training the multilayer perceptron by standard error back-propagation learning rule, the paper proposes a learning method which progressively enlarges the learning domain from a small area to the entire region. The proposed method is devised from the investigation on the roles of hidden nodes and connection weights in the multilayer perceptron which approximates a function of one variable. The validity of the proposed method was illustrated through simulations for a function of one variable and a function of two variable with many extremal points.

We have already proposed an extended fuzzy clustering algorithm which considers the importance of the data to be classified in a previous paper. In this paper, we suggest the extended fuzzy clustering algorithm based new method to slove a multiple peak detection problem, and prove experimently that this algorithm can detect the multiple peak adaptively to the noise and the shape of peaks.