Kosin Chamnongthai

The recognition of car license plate for automatic parking system

The recognition of a cars license plate for an automatic parking system is important for identifying the car at the entrance of the parking area because the car license plate has unique information for each car. This paper proposes the recognition of car license plate which is accurate and robust to environmental variation by using the cars license plate patterns according to motor vehicle regulation and a 4-layer BP neural network with supervised learning. In this method, the candidates regions of the car license plate are determined approximately according to the car license plate regulation such as color, the ratio and shape of the car license plate, the pattern of characters and numbers etc. For the results of recognition by neural networks, the candidate that has characters and numbers according to motor vehicle regulation is certified as license-plate region. Since the results of characters-pattern recognition are used to certify the license plate, the ability of license plate extraction is more accurate and the car can be identified simultaneously. The experimental results of seventy car images with the prototype of the automatic parking system show the performance of car license plate extraction rate of 96%, and the recognition rate is 92%.

Autonomous robot for a power transmission line inspection

An autonomous robot for power transmission line inspection, that can induce the voltage from the transmission line as a power source, is presented. The robot uses current transformer principle to induce the current from power transmission line and vision function to navigate. The results have shown that the robot can automatically move along the power transmission line.

Single-stage electronic ballast with class-E rectifier as power-factor corrector

A single-stage high-power-factor electronic ballast with a Class-E rectifier as a power-factor corrector is proposed. A Class-E rectifier is inserted between the front-end bridge rectifier and the bulk filter capacitor to increase the conduction angle of the bridge-rectifier diode current for obtaining low line-current harmonics. The Class-E rectifier is driven by a high-frequency sinusoidal current source, which is obtained from the square-wave output voltage of the Class-D inverter through an LC series resonant circuit. A high-frequency transformer is used for impedance matching. The experimental results for a 32-W prototype ballast are given. The switching frequency was 61.3 kHz. At full power, the power factor was 0.992 and the total ballast efficiency was 88.3%. The lamp-current crest factor was about 1.36. The simulated and experimental results were in very good agreement.

Face detection and facial feature localization without considering the appearance of image context

Face and facial feature detection plays an important role in various applications such as human computer interaction, video surveillance, face tracking, and face recognition. Efficient face and facial feature detection algorithms are required for applying to those tasks. This paper presents the algorithms for all types of face images in the presence of several image conditions. There are two main stages. In the first stage, the faces are detected from an original image by using Canny edge detection and our proposed average face templates. Second, a proposed neural visual model (NVM) is used to recognize all possibilities of facial feature positions. Input parameters are obtained from the positions of facial features and the face characteristics that are low sensitive to intensity change. Finally, to improve the results, image dilation is applied for removing some irrelevant regions. Additionally, the algorithms can be extended to rotational invariance problem by using Radon transformation to extract the main angle of the face. With more than 1000 images, the algorithms are successfully tested with various types of faces affected by intensity, occlusion, structural components, facial expression, illumination, noise, and orientation.

Face recognition system with PCA and moment invariant method

This paper proposes the integration of moment invariant and PCA for varied-pose face recognition. Firstly, the global feature is extracted by PCA for determine the minimum error. If error less than threshold, system will accepts the classification result from PCA. On the other hand, the system will rejects and moment invariant is used to analyze the local face such as nose and eyes. In the experiments, ORL face database is used to perform the method yielded the recognition rate of 96%.

Off-line signature recognition using parameterized Hough transform

This article describes a method of an off-line signature recognition by using the Hough transform to detect stroke lines from the signature image. The Hough transform is used to extract the parameterized Hough space from the signature skeleton as a unique characteristic feature of signatures. In the experiment, the backpropagation neural network is used as a tool to evaluate the performance of the proposed method. The system has been tested with 70 test signatures from different persons. The experimental results reveal a recognition rate 95.24%.

Fusion of color histogram and LBP-based features for texture image retrieval and classification

Abstract The Local Binary Pattern (LBP) operator and its variants play an important role as the image feature extractor in the textural image retrieval and classification. The LBP-based operator extracts the textural information of an image by considering the neighboring pixel values. A single or join histogram can be derived from the LBP code which can be used as an image feature descriptor in some applications. However, the LBP-based feature is not a good candidate in capturing the color information of an image, making it is less suitable for measuring the similarity of color images with rich color information. This work overcomes this problem by adding an additional color feature, namely Color Information Feature (CIF), along with the LBP-based feature in the image retrieval and classification systems. The CIF and LBP-based feature adequately represent the color and texture features. As documented in the experimental result, the hybrid CIF and LBP-based feature presents a promising result and outperforms the existing methods over several image databases. Thus, it can be a very competitive candidate in retrieval and classification application.

A match of X-ray teeth films using image processing based on special features of teeth

This paper presents a match of X-ray teeth films using image processing based on special features of teeth. This method will help the dental doctors to match simply a pair of teeth using the special features of the teeth films. Teethpsilas pictures are scanned and adjusted by a scanner and a computer, respectively, as well as then they are converted into binary code and decoded to the direction code (chain code). The chain code of each picture is compared with the statistical chain code. Therefore, the percentage of the same chain code is approximately 90% (i.e. matching same patterns) for the comparison of one root to one root (7 times) and two roots to two roots (7 times) while the percentage of the same chain code is reduced at relatively below 50% (i.e. matching different patterns) for comparison of one root to two roots (2 times).

Hand posture classification using wavelet moment invariant

In this paper, we present a wavelet moment invariants for classification a small change in rotation and subtle difference of hand posture causes misclassifying to other postures. The method combined zernike moment to capture global features and wavelet moment to differentiate between subtle variations in description can be utilized at the same time. Then, a fuzzy classification algorithm is used to classify hand posture. The classification rate obtained is 72% with of Thai sign language.

Locating essential facial features using neural visual model

Facial feature detection plays an important role in applications such as human computer interaction, video surveillance, face detection and face recognition. We propose a facial feature detection algorithm for all types of face images in the presence of several image conditions. There are two main step: the facial feature extraction from original face image, and the coverage of the features by rectangular blocks. A neural visual model (NVM) is used to recognize all possibilities of facial feature positions for the first step. Input parameters are obtained from the face characteristics and the positions of facial features not including any intensity information. For the better results, some incorrect decisions of facial feature positions are improved by image processing technique called dilation. Our algorithm is successfully tested with various types of faces which are color images, gray images, binary images, wearing the sunglasses, wearing the scarf, lighting effect, noise and blurring images, color and sketch images from animated cartoon.