Browsing by Author "Chang, JY"

Now showing 1 - 18 of 18

2-D discrete signal interpolation and its image resampling application using fuzzy rule-based inference
(2000-09-01)
Chen, JL
;
Chang, JY
;
Shieh, KL
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper describes an interpolation algorithm for two-dimensional (2-D) discrete signals using fuzzy rule-based inference. The original signal is estimated by the main-surface function in the interpolation region, and four sub-plane functions surrounding the interpolation region. The main-surface is a bilinearly interpolated function passing through four signal samples in the interpolation region and the four sub-planes reflect the tendencies of pixels from the left, right, up, and down of the interpolation region. Drawing fuzzy inferences about signals from these five functions, we can estimate original signals very well even when the signals are buried in noise. We verified the method by computer simulations of some assumed 2-D signals and by resampling of the actual image data. (C) 2000 Elsevier Science B.V. All rights reserved.
A new scene analysis using genetic algorithm based fuzzy ID3 method
(2005)
Chang, JY
;
Cho, CW
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, we utilize a neural network based machine learning algorithm to segment natural objects in outdoor scene images. We have developed a genetic algorithm based fuzzy ID3 method, which can build a fuzzy decision tree to summarize the regularities existing in the data set. Using this method, we then propose a road scene analysis system, by which natural element segmentation rules can be learned from several road scene images. In the image analysis phase, the natural element regions are obtained through inference on these learned rules. Moreover, we can apply image groundtruthing to further improve the classification accuracy. The testing results have demonstrated that the object segmentation accuracy is quite high.
A novel prosodic-information synthesizer based on recurrent fuzzy neural network for the Chinese TTS system
(2004-02-01)
Lin, CT
;
Wu, RC
;
Chang, JY
;
Liang, SF
;
電控工程研究所
In this paper, a new technique for the Chinese text-to-speech (TTS) system is proposed. Our major effort focuses on the prosodic information generation. New methodologies for constructing fuzzy rules in a prosodic model simulating human's pronouncing rules are developed. The proposed Recurrent Fuzzy Neural Network (RFNN) is a multilayer recurrent neural network (RNN) which integrates a Self-constructing Neural Fuzzy Inference Network (SONFIN) into a recurrent connectionist structure. The RFNN can be functionally divided into two,parts. The first part adopts the SONFIN as a prosodic model to explore the relationship between high-level linguistic features and prosodic information based on fuzzy inference rules. As compared to conventional neural networks, the SONFIN can always construct itself with an economic network size in high learning speed. The second part employs a five-layer network to generate all prosodic parameters by directly using the prosodic fuzzy rules inferred from the first part as well as other important features of syllables. The TTS system combined with the proposed method can behave not only sandhi rules but also the other prosodic phenomena existing in the traditional TTS systems. Moreover, the proposed scheme can even find out some new rules about prosodic phrase structure. The performance of the proposed RFNN-based prosodic model is verified by imbedding it into a Chinese TTS system with a Chinese monosyllable database based on the time-domain pitch synchronous overlap add (TD-PSOLA) method. Our experimental results show that the proposed RFNN can generate proper prosodic parameters including pitch means, pitch shapes, maximum energy levels, syllable duration, and pause duration. Some synthetic sounds are on-line available for demonstration.
Anisotropic self-diffraction under Bragg mismatching
(2001-05-01)
Wang, B
;
Sun, CC
;
Yang, TH
;
Ye, P
;
Chang, JY
Anisotropic self-diffraction (ASD) under Bragg mismatch has been studied. We derive a solution that can describe well the diffraction characteristics of the anisotropic self-diffraction under Bragg mismatch. The solution is useful for estimating the Bragg constraint when the ASD is applied to optical information processing. Both the theory and the experiment are presented.
Applying fuzzy logic in the modified single-layer perceptron image segmentation network
(2000-03-01)
Chang, JY
;
Chen, JL
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, we propose a fuzzy-logic-based modified single layer perceptron (MSLP) image segmentation network for object extraction. We select a sigmoid gray level transfer function with the aid of the input image histogram and map the input gray levels into the interval [0,1]. Then we adopt the linear index of fuzziness of the output nodes as the error function of the image segmentation system to incorporate the learning capability of a neural network. Our scheme can successfully extract objects from the background. To further enhance the capability of the segmentation system, the proposed network is incorporated with fuzzy if-then rules to adaptively adjust the threshold of the activation function of the MSLP output neuron for best matching the local characteristics of the image. Fuzzy if-then rules involving the edge intensities and vertical positions of pixels are reasoned to determine the threshold adaptively. From the results of segmenting forward looking infrared (FLIR) images, better segmentation images have been obtained by incorporating fuzzy if-then rules with the MSLP segmentation technique. As demonstrated by this study, it is promising and worthy of study that incorporating human knowledge in terms of Fuzzy rules into a designed numerical algorithm can further improve performance, not only in the segmentation problem we present.
Automated facial expression recognition system using neural networks
(2001-05-01)
Chang, JY
;
Chen, JL
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper proposes an automated facial expression recognition system using neural network classifiers. First, we use the rough contour. estimation routine, mathematical morphology, and point contour detection method to extract the precise contours of the eyebrows, eyes, and mouth of a face image. Then we define 30 facial characteristic points to describe the position and shape of these three facial features. Facial expressions can be described by combining different action units, which are specified by the basic muscle movements of a human face. We choose six main action units, composed of facial characteristic point movements, as the input vectors of two different neural network-based expression classifiers including a radial basis function network and a multilayer perceptron network. Using these two networks, we have obtained recognition rates as high as 92.1 % in categorizing the facial expressions neutral, anger, or happiness. Simulation results by the computer demonstrate that computers are capable of extracting high-level or abstract information like humans.
Classifier-augmented median filters for image restoration
(2004-04-01)
Chang, JY
;
Chen, JL
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Developed in this paper is a new approach that augments a fuzzy classifier to determine whether or not the operating pixel, centered in the sliding window, should be involved with the impulse noise filtering process. Owing to the inclusion of the fuzzy K-nearest neighbor (K-NN) scheme, any central operating pixel that is not noise corrupted can be effectively detected and then left unchanged. Thus, the unnecessary pixel replacement can be avoided and the details and signal structure of the image will be best retained. If the center point is found to be noise corrupted, the proposed classifier-augmented median filter facilitates the filtering action only on a subset of pixels, which are not noise contaminated in the window. Due to this impulse pixel exclusion, the biased estimation caused from impulses can be eliminated and, thus, obtains a better estimation of the center pixel. Experimental results showed that this new approach largely outperformed several existing schemes for image noise removal.
Digital image translational and rotational motion stabilization using optical flow technique
(2002-02-01)
Chang, JY
;
Hu, WF
;
Cheng, MH
;
Chang, BS
;
電控工程研究所
This paper proposes a new digital image stabilization (DIS) systems based on optical flow technique. Unlike previous DIS systems developed mainly for removing the translational motion disturbance, the proposed system removes not only the translational but also the rotational motion disturbances. A computational scheme that facilitates the local motion vector field to estimate the global disturbing translational and rotational motions is developed. First, the optical flow technique is used to estimate the local motion vector field of the image, yielding the velocity of each pixel in the current image frame. Then, the global translational and rotational motion parameters are determined in terms of the least squares estimation. Finally, these motion vectors are used to generate the counterbalance signals for removing the disturbance motion. Owing to the additional ability of the rotational motion removal, the new DIS system suppresses the undesirable translational and rotational disturbances effectively and thus enhances the DIS performance significantly.
Fuzzy perceptron learning and its application to classifiers with numerical data and linguistic knowledge
(1995)
Chen, JL
;
Chang, JY
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Fuzzy perceptron neural networks for classifiers with numerical data and linguistic rules as inputs
(2000-12-01)
Chen, JL
;
Chang, JY
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents a novel learning algorithm of fuzzy perceptron neural networks (FPNNs) for classifiers that utilize expert knowledge represented by fuzzy IF-THEN rules as well as numerical data as inputs. The conventional linear perceptron network is extended to a second-order one, which is much more flexible for defining a discriminant function. In order to handle fuzzy numbers in neural networks, level sets of fuzzy input vectors are incorporated into perceptron neural learning, At different levels of the input fuzzy numbers, updating the weight vector depends on the minimum of the output of the fuzzy perceptron neural network and the corresponding nonfuzzy target output that indicates the correct class of the fuzzy input vector, This minimum is computed efficiently by employing the modified vertex method to lessen the computational load and the training time required. Moreover, the; pocket algorithm, called fuzzy pocket algorithm, is introduced into our fuzzy perceptron learning scheme to solve the nonseparable problems, Simulation results demonstrate the effectiveness of the proposed FPNN model.
Fuzzy stroke type identification for on-line Chinese character recognition
(1998)
Chang, JY
;
Wan, MH
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents an on-line Chinese character stroke type recognition system based on fuzzy set theory. According to the writing strobe sequence, each character is described by an 1-D stroke string model. The input written characters can be loosely constraints, which are quite flexible on size and allow various variations. The stroke segments of input strokes are extracted firstly and the strokes of input character are then identified as a sequence of primitive stroke types by a fuzzy methodology. A character recognition system, based on this stroke type identification scheme and Modified Dynamic Programming Forward (MDPF) matching, Modified Dynamic Programming Backward (MPDB) matching, and A* matching algorithms, has been built and tested to be very successful.
Fuzzy-logic-based modified single-layer perceptron segmentation network
(1998)
Chang, JY
;
Chen, JL
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, we propose a modified singlelayer perceptron (MSLP) segmentation network for object extraction. We select a sigmoid gray level transfer function from the histogram of the input image and map the input gray levels into the interval [0, 1]. Then we adopt the linear index of fuzziness of the output nodes as the error function of the image segmentation system to incorporate the learning capability of a neural network. Our scheme can successfully extract multiple objects with different gray levels. To further enhance the capability of the segmentation system, the proposed network is incorporated with fuzzy if-then rules to adaptively adjust the threshold of the activation function of the output neuron for best matching the local characteristics of the image. Fuzzy if-then rules involving the edge intensities and vertical positions of pixels are reasoned to determine the threshold adaptively. From the result of segmenting the forward looking infrared (FLIR) image, a better segmentation image has been obtained by incorporating fuzzy if-then rules with the MSLP segmentation technique.
Genetic algorithm based fuzzy ID3 algorithm
(2004)
Chang, JY
;
Cho, CW
;
Hsieh, SH
;
Chen, ST
;
電控工程研究所
In this paper, we propose a genetic algorithm (GA) based fuzzy ID3 algorithm to construct a fuzzy classification system with both high classification accuracy and compact rule base size. This goal is achieved by two key steps. First, we optimize by GA the parameters controlling the means and variances of fuzzy membership functions and leaf node conditions for tree construction. Second, we prune the rules of the tree constructed by evaluating the effectiveness of the rule, and the remaining rules are retrained by the same GA proposed. Our proposed scheme is tested on various famous data sets, and its results is compared with C4.5 and IRID3. Simulation result shows that our proposed scheme leads to not only better classification accuracy but also smaller size of rule base.
Genetic algorithm-based neural fuzzy decision tree for mixed scheduling in ATM networks
(2002-12-01)
Lin, CT
;
Chung, IF
;
Pu, HC
;
Lee, TH
;
Chang, JY
Future broad-band integrated services networks based on the asynchronous transfer mode (ATM), technology are expected to support multiple types of multimedia information with diverse statistical characteristics and quality of service (QoS) requirements. To meet these requirements, efficient scheduling methods are important for traffic control in the ATM networks. Among the general scheduling schemes, the rate monotonic algorithm is simple enough to be used in high-speed networks, but it does not attain as high a system utilization as the deadline driven algorithm does. However, the deadline driven scheme is computationally complex and hard to implement in hardware. The mixed scheduling algorithm is the combination of the rate monotonic algorithm and the deadline driven algorithm; thus it can provide most of the benefits of these two algorithms. In this paper, we use the mixed scheduling algorithm to achieve high system utilization under the hardware constraint. Because there is no analytic method for the schedulability test of the mixed scheduling, we propose a genetic algorithm-based neural fuzzy decision tree (GANFDT) to realize it in a real-time environment. The GANFDT combines the GA and a neural fuzzy network into a binary classification tree. This approach also exploits the power of the classification tree. Simulation results show that the GANFDT provides an efficient way to carry out the mixed scheduling in the ATM networks.
Image blocking artifact suppression by the modified fuzzy rule based filter
(2003)
Chang, JY
;
Lu, SM
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper we propose the modified fuzzy rule-based (MFRB) filter for JPEG-encoded image blocking noise removal. The proposed filter generalizes the FRB filter which is a weighted average of the center pixel and its neighboring pixels in a working window. The weighting coefficients of the MFRB filter depend on the gray level difference, the spatial distance and orientation, and the variance of the pixels in the window. By the least mean square (LMS) learning over representative JPEG-encoded images, the best membership function, i.e., the weight coefficients, of the MFRB filter for blocking noise suppression can be determined. By the simulation results, our proposed scheme has demonstrated the effectiveness and robustness in blocking noise removal. Compared to several post-filtering approaches, our MFRB filter has produced the best SNR improvement among the resultant images. This objective superiority is also confirmed visually by the sharper and more pleasing image obtained.
Measurements of partition coefficients in water-1,1,1,2-tetrafluoroethane by countercurrent chromatography
(2001)
Yu, T
;
Chang, JY
;
應用化學系
;
Department of Applied Chemistry
Recently, refrigerant 1, 1, 1, 2 -tetrafluoro ethane (R134a) was considered as an extraction solvent alternative to supercritical fluid CO2. We measured water-liquid R134a partition coefficients of some organic compounds using a high-speed countercurrent chromatography (HSCCC). The experiment could be operated by moderate modification of the commercial instrument. The partition coefficient data were obtained under 15 degrees, 25 degrees, and 45 degreesC.
Scene analysis system using a combined fuzzy logic-based technique
(2002-05-01)
Chang, JY
;
Cho, CW
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper proposes a two-stage scene analysis scheme using a combined fuzzy logic-based technique. The first stage begins with generating fuzzy rules to describe the scene. Based on these fuzzy classification rules. each image pixel is inferred and then classified to the natural object category with the largest membership degree. The second stage involves a newly derived fuzzy K-nearest neighbor algorithm that further refines the classification result obtained. With this second stage, the proposed system is robust because it is demonstrated to be insensitive to the variations of membership functions and image noise contamination. Simulations of real world images have shown that the proposed scheme is very successful and the results are visually confirmed by human observation. The satisfactory results achieved in this paper suggest the feasibility of developing similar systems for other types of images aiming at image description problems.
Second-order asymmetric BAM design with a maximal basin of attraction
(2003-07-01)
Chang, JY
;
Cho, CW
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Bidirectional associative memory (BAM) generalizes the associative memory (AM) to be capable of performing two-way recalling of pattern pairs. Asymmetric bidirectional associative memory (ABAM) is a variant of BAM relaxed with connection-weight symmetry restriction and enjoys a much better performance than a conventional BAM structure. Higher-Order associative memories (HOAMs) are reputed for their higher memory capacity than the first-order counterparts, yet there are few HOAMs design schemes proposed up to date. To this end, we are concerned in this paper with designing a second-order asymmetric bidirectional associative memory (SOABAM) with a maximal basin of attraction, whose extension to a HOABAM is possible and straightforward. First, a necessary and sufficient condition is derived for the connection weight matrix of SOABAM that can guarantee the recall of all prototype pattern pairs. To respect the complete recall theorem, an adaptive local training rule, which is adaptive in the learning step size and updates only the entries in the connection weight related to the most needful bit of a prototype, is formulated and it leads to better results and faster design. Then derived is a theorem, designing a SOABAM further enlarging the quantities required to meet the complete recall theorem will enhance the capability of evolving a noisy pattern to converge to its association pattern vector without error. Based on this theorem, our algorithm is also modified to ensure each training pattern is stored with a basin of attraction as large as possible. Computer simulations over the color graphics adapter (CGA) fonts have demonstrated the superiority of the proposed local training rule over other prevailing BAM schemes.