Tsun-Wei Chang

A BACK PROPAGATION BASED REAL-TIME LICENSE PLATE RECOGNITION SYSTEM

License plate recognition systems have been used extensively for many applications including parking lot management, tollgate monitoring, and for the investigation of stolen vehicles. Most researches focus on static systems, which require a clear and level image to be taken of the license plate. However, the acquisition of images that can be successfully analyzed relies on both the location and movement of the target vehicle and the clarity of the environment. Moreover, only few studies have addressed the problems associated with instant car image processing. In view of these problems, a real-time license plate recognition system is proposed that recognizes the video frames taken from existing surveillance cameras. The proposed system finds the location of the license plate using projection analysis, and the characters are identified using a back propagation neural network. The strategy achieves a recognition rate of 85.8% and almost 100% after the neural network has been retrained using the erroneously recognized characters, respectively.

A fuzzy feature clustering with relevance feedback approach to content-based image retrieval

The increasing number of digitized images required an efficient image retrieval system. In this paper, we demonstrate the fundamental principles, implementation methods, performance evaluations, and experimental results from the proposed model. We present a region-based prototype image retrieval system named FuzzyImage. The system is characterized by feature vectors. First, we segment an image into regions depending on clustering similar feature vectors by fuzzy c-means. Next, a similar measurement is used to evaluate the similarity between the query image and incorporated regions. The users can select the most interesting regions from 5 sample images that pop-up, and by feedback to the system. Based on the selected individual regions of query images, the overall similarity helps filter out irrelevant images in a database after relevance feedback and enables a simple user-oriented query interface for a region-based image retrieval system. This algorithm is implemented and tested on general-purpose images. This project makes three main contributions to a region-based CBIR system. First, a region segmentation method is employed in the FuzzyImage system. Second, this system takes the users intuition into consideration and designs a user-oriented interface to directly search the database. Thirdly, we evaluate retrieval precision of the system to support this theoretical claim.

Efficient entropy-based features selection for image retrieval

Information retrieval systems should provide users quick access to desired information. There are no established ways for inexperienced users to explicitly express queries for retrieving images from ecological databases. This study proposes an entropy-based feature selection strategy for finding images of interest from databases. Six visual features are used to represent birds, and hence used to formulate search queries. The proposed method is tested on a real world bird database and the experimental results demonstrate the effectiveness of the presented work.

A Ubiquitous Interactive Museum Guide

This paper proposes a PDA-based guide and recommendation system. The intelligent guide provides an alternative to the current audio-based guide tools. Techniques from both data mining and extension theory and RFID technology are deployed to provide better interaction between exhibitions and visitors.

An ontology oriented region-based image retrieval strategy

A novel and more effective region-based image retrieval strategy is presented based on semantic ontology. An unsupervised segmentation algorithm splits images into regions that are subsequently used as basis by the ontology-based strategy. The approach comprises three stages, namely automatic region generation, categorization and ontology construction. When receiving a query for a specific object, the search engine will, in addition to conventionally matched images, also find candidates through the semantic ontology using low level features. The proposed approach can thus find a richer set of related candidate images than traditional image retrieval approaches. This strategy is particularly useful for vague queries encountered by inexperienced users that are not trained in searching for images by the means of low-level features. The experimental results demonstrate the effectiveness of the proposed approach.

An Interactive Handheld Device-Based Guide System Using Innovative Techniques

The traditional guide systems only provide limited information for users. Contrasting to the passive guide systems, an interactive guide and recommendation system is proposed in this paper. Integrating state-of-the-art technologies, including RFID and association rules, the proposed system offers flexible strategies to enhance the guide interaction. With the characteristics of light and portability, the handheld devices can perfectly realize an application of guide and recommendation system for relics, cultures, paintings, etc. The design methodology is illustrated and the experimental results are demonstrated in this paper.

A fuzzy inference model for image segmentation

We present a novel method to segment objects in images based on the similarity measurement of fuzzy gray level technique in this paper. In our model, we classify the processing steps into three stages. First, we utilize the attributes of luminance and chromaticity components of HLS color coordinate system to form a fuzzy gray level. These attributes can describe the relationship between different frequent colors and the image can be transferred to smooth gray level, which can capture the objects in images. Second, we reduce the gray levels of image pixels to lower gray levels to speed up computation. Third, we label each root pixel based on a similarity measurement. We perform a sliding window to move from one block to the next one. The similarity of the two root pixels blocked by the sliding window depends on their neighboring pixels. Via the similarity computation, we assign a label number to the root pixels. We generate objects from grouping different labels. The image data are classified by fuzzy gray level technique and the objects are segmented from images. According to the simulation results, our model shows the efficiency and effectiveness for image segmentation.

Designing a Shape-Based Image Retrieval System by Grey Relational Analysis and GM(1,N) Model

With the explosive growth of multimedia applications, the quality and ability to retrieve images in an efficient way is a challenge to researchers. A new shape-based image retrieval model is presented to improve the recall rate of image retrieval system. The retrieval procedures consist of two major steps: the feature representation of objects shape and the image retrieval method. The shape signatures are extracted along the object boundary to form a data sequence for grey model. After deriving the feature sequences, both grey relational analysis and GM(1,N) model are integrated to construct the proposed image retrieval model. A fish data set is selected to test the reconstructed error rates when different comparative images are chosen for the GM(1,N) model. Experimental results from the fish shape library verify the effectiveness of the proposed model.

Improving image retrieval efficiency using a fuzzy inference model and genetic algorithm

The typical approaches for content based image retrieval extract signatures such as color, shape, and texture from each image and map the features into a d-dimensional metric space. According to the similarity measurement, the images closer to the signatures are shown to the users and constitute the query results. Such a retrieval method is prohibited in matching similar images from global similarity measurement that fails to consider partially matched images. In this paper, a genetic algorithm based approach is proposed for the retrieval of images which may have been rotated and transposed. The approach utilizes fuzzy inference model to segment an image into some regions according to imagery contents. Then, the color histograms are calculated from each region to reflect content feature property. After users submit a query image, the system can generate a population pool as the potential solutions from possible combinations of query image regions. The system applies the genetic operations and maintains the best top 5 images. Rotation and position displacement of objects are taken into account. Unlike retrieval based on individual region matching, the proposed scheme performs an overall similarity computation. A thorough experiment demonstrates the robustness of the proposed system.

Query-by-Humming/Singing of MIDI and Audio Files by Fuzzy Inference System

Music Information Retrieval (MIR) is a crucial topic in the domain of information retrieval. According to major characteristics of music, Query-by-Humming system retrieves interesting music by finding melody that contains similar or equal melody to the humming query. Basing on the designed fuzzy inference model a novel Query-by-Humming/Singing system is proposed to extract pitch contour information from WAV and MIDI files in this paper. To verify the effectiveness of the presented work, the MIREX QBSH Database is employed as our experimental database and a large amount of human vocal data is used as query to test the robustness of MIR. Then, the Longest Common Subsequence (LCS) is served as an approximate matching algorithm to identify the most related top 5 music as an evaluation standard for the system. Experimental results show that the proposed system achieves 85% accuracy in the top 5 retrievals.