Applied Mechanics and Materials Vols. 373-375

Abstract: Object detection is the important foundation of visual tracking. In this paper, a real-time object detection algorithm based on back-projection was presented. Firstly, according to the principle of back-projection, the objects probability image is calculated by objects color histogram model, and then we determine the object on the basis of some contour strategy in that image. Experimental results show that the proposed algorithm accurately detected the position of object in real-time if the contour of object change within a certain range and the color of object is distinct.

Abstract: Bar code technology has been widely used with the development of computer application, which studies information identification by bar code and conversion from bar code information to computer language for automatic input, identification and statistics. This paper presents system design and realization of recognition system for one-dimensional bar code according with ISBN encoding rule. Encoding rule of common bar code is introduced at first. Bar code image is preprocessed by binaryzation, then the image noise caused by print quality and dirt is filtered by a mean value method. The actual width of bars and intervals are measured to identify contained digital information in one-dimensional bar code image. The result show that the method presented has good accuracy, quick speed and strong robustness for realtime application.

Abstract: In this paper, the average intensity and the effect of focus on it for disyllabic words in Chinese is analyzed. It is found that, the average intensity of rhyme in the first syllable is comparatively great, while for the onset, due to intervocalic voicing, that in the second syllable is great. Focus has great effect on intensity. Analysis shows that, compared to the onset and the first syllable, the effect on the rhyme and on the second syllable is great.

Abstract: Structured light extraction is an important problem in 3-D measurement of structured light. Due to traditional structured light extraction methods are mostly applied in small visual field and simple environment. Therefore, a novel method is proposed based on the feature of structured light aiming at the large field and complex surroundings. Firstly, a region of interest (ROI) is obtained through a third sensor in order to reduce calculations and process conveniently. And then, structured light is extracted according to its feature of brightness distribution module and geometrical morphology trait. Experiment shows that this new method not only extracts structured light in such large field with strong background noise well but also reduces the computation to meet real-time requirement.

Abstract: An appearance-based feature set is proposed. With Hidden Markov Model (HMM) handling any temporal variance, the contributions of features, which are from full foreground sequence and from temporal difference sequence, are compared in details by methods which are based on feature selecting and feature voting. The experimental analysis shows that the comparative contributions can be achieved for human action identifying by the two data sources. This introduces the opportunity to analyze human behavior based on temporal difference sequence instead of full foreground sequence, and validates the far-reaching significance of this work.

Abstract: The paper analyzes the working theory of a Speech Conversion System from PuTongHua to Cantonese based on iFLY MSP 2.0. In the system, QISR interface is chosen to complete speech recognition function which is the key technology to convert the voice information into the corresponding text information. Moreover, the QTTS interface is chosen to complete the text to speech function which is the key technology to transform the text which is the result of the speech recognition into the spoken information in Cantonese and then output. Finally, the computer assisted learning system is designed successfully in the environment of Visual C++ 6.0.

Abstract: It is difficult to detect the edges of objects in side scan sonar images due to the complex background, bad contrast and deteriorate edges. Therefore, it is important to remove noise from side scan sonar images. The traditional de-noising methods for optical images may not work well on the sonar image. In this paper, an adaptive de-noising approach is used. The side scan sonar image is first filtered using mean filter to remove the rough noise, then a weighted function is generated using spatial distance filter and intensity distance filter. The parameters are adaptive according to the sonar image. The experimental results indicate that it is an effective de-noising method for underwater sonar image.

Abstract: We propose a new perspective on image encryption using chaotic signal and blind source separation. The original image is permuted by the chaotic signal and then mixed with key images. In the receiver, blind source separation technique is used to separate the components of the original image from the ciphertexts. Then chaotic signal is again used to restore the pixels to recover the original image. The experimental results demonstrate that the key space is large enough to resist the brute force attack and the distribution of gray values of the encrypted image has a random-like behavior.

Abstract: Image edge detection technology which is applied to the hydrophobic grade discrimination of composite insulator is of great significance. How to accurately detect the edge of the water droplets is very important in the process of automatic identification of composite insulator hydrophobicity. This paper detects composite insulator hydrophobic image edge with image edge detection technology. Firstly, adaptive histogram equalization technique is used to enhance image, reducing the similarity of target and background which is caused by the transparency of water. Then the paper uses the improved Canny algorithm which is introduced Ostu threshold method to detect the image edge. The experimental results show that the improved algorithm can accurately acquire the water droplets edge of composite insulator surface, which can lay a foundation for automatic discrimination of composite insulator hydrophobic grade.

Abstract: This paper presents a medical bottle detecting system based on machine vision, and completing the tightness detection task for the installation of vial and cap. The first step is experimental platform building; secondly, using MATLAB software for operations such as histogram equalization, image enhancement, image segmentation and boundary signature; finally, finalizing the design of the system software with MATLAB GUI. The systems can significantly reduce testing cost, improve detection quality, speed up the system efficiency and realize non-contact real-time detection in industrial field.