Supot Nitsuwat

Prediction of stock price using an adaptive Neuro-Fuzzy Inference System trained by Firefly Algorithm

The substance of the design of Adaptive Neuro-Fuzzy Inference System (ANFIS) can be seen as an optimization problem to find the best parameters with minimal error function. This paper proposes a combination of the Firefly Algorithm and Adaptive Neuro-Fuzzy Inference System. The fuzzy neural network model will be trained by the Firefly Algorithm, and applied to predict stock prices in the Vietnam Stock Market. The experiments will compare performance between the proposed system and ANFIS trained by the Hybrid Algorithm, Back Propagation and Particle Swarm Optimization (PSO). The experimental results show that the system has reasonable efficient performance.

Traffic Sign Recognition System for Roadside Images in Poor Condition

Traffic sign detection and recognition is a difficult task, especially if we aim at detecting and recognizing signs in images captured under poor conditions. Complex backgrounds, obstructing objects, inappropriate distance of signs, shadow, and other lighting-related problems may make it difficult to detect and recognize signs in both rural and urban areas. In this paper we propose and test a system that employs image pre-processing, color filtering, color segmentation for traffic sign detection at the detection stage, feature extraction and trained neural networks for unique identification of signs at the recognition stage. The traffic sign detection and recognition system has been tested on actual roadside images taken under poor conditions. The images were selected in order to test the efficiency of the system under challenging conditions of inappropriate distance, traffic sign size, poor lighting and complex background. Suggestions are made for improving the performance of the system.

Traffic sign recognition by color segmentation and neural network

An algorithm is proposed for traffic sign detection and identification based on color filtering, color segmentation and neural networks. Traffic signs in Thailand are classified by color into four types: namely, prohibitory signs (red or blue), general warning signs (yellow) and construction area warning signs (amber). A color filtering method is first used to detect traffic signs and classify them by type. Then color segmentation methods adapted for each color type are used to extract inner features, e.g., arrows, bars etc. Finally, neural networks trained to recognize signs in each color type are used to identify any given traffic sign. Experiments show that the algorithm can improve the accuracy of traffic sign detection and recognition for the traffic signs used in Thailand.

An improved local binary pattern for edge detection of images

This research proposed the improved H-LBP for edge detection, TH-LBP. It applied the Hyperbolic Tanh function instead of the H function. In this method, the continuous and thick edges can be obtained. In addition, the method parameters could be adjusted to increase the edge details in the low contrast area. This method can solve the problem in the case that the two different local areas obtain the same result.

Application of logarithm, wavelet and contourlet transform for video-frame quality improvment

Video shot boundary detection is important step for the research in the content analysis and retrieval fields. In this paper, firstly we presented efficient method for Video-frame quality improvement to suppress flash occurred within video frame using logarithm, wavelet and contourlet transform. In addition, wavelet and contourlet transform also performed denoising. Secondly, for shot boundary detection, we used gray-scale histogram differences with edge change ratio. Furthermore, the adaptive threshold algorithm for shot transition detection was proposed. The experiment results showed that using logarithm with contourlet transform gain the precision and recall higher than using logarithm with wavelet transform.