Fumio Itami

A realization of digital filter banks for reconstruction of uniformly sampled signals from nonuniform samples

This paper proposes a filter bank structure for the reconstruction of uniformly sampled signals from nonuniform samples, on the basis of a conventional filter bank framework. First, a digital representation of filters used in the filter bank is derived. Next, an efficient filter bank structure is also provided. A proposed scheme includes DFT modulation of a fixed prototype filter and simple fractional delay elements, so that its design cost is low. Finally, simulation results are shown in which we confirm that uniform samples are reconstructed by the proposed scheme.

A design of cyclic filter banks

Cyclic filter banks are appropriate for processing finite-length signals such as images since cyclic convolutions are used to process them. So far, it has been shown that the two channel linear phase orthogonal filter banks, which can not be realized as non-cyclic LTI systems, can be realized as cyclic LTI systems. This paper proposes a design of the cyclic filter banks with DFT modulations. First, we show the polyphase representations and perfect reconstruction conditions. Next, we discuss the advantages of the proposed structure by studying the essential property of the cyclic LTI systems. Finally, we give a design example of the proposed filter banks, and show the reconstructed images processed by the proposed filter banks.

Multirate Filter Bank-based Conversion of Image Resolution

This paper proposes a multirate processing scheme for the conversion of image resolution by using filter banks. A resolution-converted signal is synthesized from decomposed ones by the filter banks. A new design of filter banks for resolution conversion is proposed. The design strategy is that we make the output of synthesis filter banks nearly equal to the one of a direct resolution conversion approach. Simulation results are also given, in which we examine the performance of the proposed synthesis scheme

A Change Detection Method for Image Sequences Based on Physical Models

Change detection methods detect changes between two frames in an image sequence. Fundamental technique for detecting changes uses the difference image between two frames. A change of each pixel is detected if a difference value exceeds pre-set thresholds, which is determined based on an estimated value of variance of noise on images. Not only the noise on images but also illumination changes between frames is one of critical problems for change detection. A recently proposed approach shows better results by using thresholds derived from average of difference image over areas which are estimated as non-change parts. However, such a threshold may not be appropriate since this approach uses no physical parameters such as light sources, reflection of objects. This paper proposes a new change detection method based on physical model, which describes physical parameters such as light sources and reflection of objects, known as illumination model. First, we show derivation of threshold based on illumination model. Threshold is derived from angle of light of sources, gray level of background objects, and normal-vector of background objects. Next, change detection algorithm using such a threshold is shown. Finally, we show experimental results and comparison, in which proposed method improves accuracy of detection results, compared to the conventional method.

A study on digital filter banks for reconstruction of uniformly sampled signals from nonuniform samples

This paper provides a study on a practical digital filter bank based on a conventional filter bank framework for the reconstruction of uniformly sampled signals from nonuniform samples. The digital filter bank structure is studied in an implementation and design point of view, and its reconstruction performance is compared with a least squares scheme.

A design of a synthesis filter bank with fractional scalability factors

This paper provides a simple design procedure of a synthesis filter bank which synthesizes the decomposed signals of an analysis filter bank to obtain a rescaled signal. The design procedure is derived from the discussion on the equations which are based on the relation between the input and output of the filter bank. It is demonstrated through simulation results that its design procedure is simplified compared with that of a conventional approach.

A change detection method for image sequences based on physical models

Change detection methods detect changes between two frames in an image sequence. Fundamental technique for detecting changes uses the difference image between two frames. A change of each pixel is detected if a difference value exceeds pre-set thresholds, which is determined based on an estimated value of variance of noise on images. Not only the noise on images but also illumination changes between frames is one of critical problems for change detection. A recently proposed approach shows better results by using thresholds derived from average of difference image over areas which are estimated as non-change parts. However, such a threshold may not be appropriate since this approach uses no physical parameters such as light sources, reflection of objects. This paper proposes a new change detection method based on physical model, which describes physical parameters such as light sources and reflection of objects, known as illumination model. First, we show derivation of threshold based on illumination model. Threshold is derived from angle of light of sources, gray level of background objects, and normal-vector of background objects. Next, change detection algorithm using such a threshold is shown. Finally, we show experimental results and comparison, in which proposed method improves accuracy of detection results, compared to the conventional method.

A design method of multi-channel cyclic filter banks with less multiplications by genetic algorithm

This paper proposes a design method of multi-channel cyclic filter banks with DFT modulation by genetic algorithms. First, their polyphase representations and perfect-reconstruction conditions are given. Next, based on such structure, we show optimization procedure using genetic algorithms, which enable us to realize DFT modulated cyclic filter banks with prototype filters that have coefficients of the sum of power-of-two numbers. Finally, we give simulation results of the algorithms