Chuan-Pin Lu

A histogram based data-reducing algorithm for the fixed-point independent component analysis

This paper proposes a histogram based data-reducing algorithm for improving the performance of the fixed-point independent component analysis (FastICA). This data-reducing independent component analysis (DR-FastICA) is based upon two statistical criteria to keep the histogram contour of processed data. This algorithm uses two steps (a coarse step for data sampling and a fine one for data tuning) to improve the performance of FastICA. Experimental results show that the proposed algorithm can reduce the computation time and implementation memory needed for executing FastICA, especially for large amounts of data (e.g. 1024x1024 images).

Recognition of the Ambulance Siren Sound in Taiwan by the Longest Common Subsequence

This paper proposes a method to recognize the ambulance siren sound in Taiwan. Since the ambulance siren sound is composed by a high frequency and a low frequency signal, the sequence of the frequency change is used to be the feature of the sound. The input sound is divided into frames, and each frame is classified into high frequency or low frequency. The Longest Common Subsequence (LCS) is used to compare the arrangement of the frequencies in the frames. We use the results of LCS to recognize whether the sound comes from an ambulance. Real sounds are used to show the performance. According to the experimental results, the accuracy rate is 85%.

Geometric transformation based independent component analysis for mixed image separation

Abstract An independent component analysis (ICA) method for image separation by geometric transformation of a scatter diagram is proposed. Geometric transformation and normalization are used to project mixed image signals to independent component space. This method includes four procedures: data correction, whitening, geometric rotation, and slant compensation. Several synthetic mixed image and real applications are used to evaluate the performance of the proposed method. From experimental results, mixed images are separated accurately by the proposed method.

Mobile Care Robot Localization and Guidance Using Image Structure Map

This paper presents a localization and guidance method for mobile robots based on computer vision. The purpose of this study was to use CCTV-system video images to position and guide mobile robots. This method enables direct observation of the robot and is not influenced by the characteristics of a specific region or area, changes in the environment at a certain time are used to guide the robots movement. In this study, an image structure map was designed to guide the movement of the robot, and motion detection was integrated to position the robot. Throughout the movement process, ultrasound was used to detect obstacles and send feedback to the image structure map, which then immediately labeled the map and changed the course of the robot. This method may be directly employed for care robots and can be extended to other applications in the future.