Kota Irie

Construction of an intelligent room based on gesture recognition: operation of electric appliances with hand gestures

This paper proposes an intelligent room that is free of operators position based on gesture recognition technologies. Intention and position of an operator are recognized by detecting hand waving, and pan-tilt cameras are zoomed and focused on the operator. The hand region is extracted using color information, and direction or number of fingers and motion of the hand region are detected. Home appliances such as a television set are controlled by using the direction or number of fingers and hand motions.

Gesture recognition based on the detection of periodic motion

In this paper, we propose a method to recognize periodic gestures from images. The proposed method uses a amplitude spectrum and a phase spectrum that are obtained by applying Fast Fourier Transform (FFT) to a time series of intensity images. FFT is applied to each pixel of low-resolution images. The method consists of 2 steps. First, the method detects p eriodic motion regions from the amplitude spectrum. Secondly, the method uses the phase spectrum in the detected periodic motion region to classify the gestures. The proposed method is robust to lighting conditions and individual differences in skin color because it does not rely on color information. Several experiments are performed to demonstrate the effectiveness of the proposed method.

3D Measurement by Distributed Camera System for Constructing an Intelligent Room

We are constructing an intelligent room in which an operator can control home appliances such as a TV set with intuitive gestures. In this paper, we discuss three dimensional (3D) measurement of the intelligent room using a distributed camera system. The room carries CCD cameras with pan, tilt, and zoom functions. An operator makes hand waving and the room detects it. 3D position of the waving hand is measured using the cameras, and then pan, tilt and zooming-up of the cameras are carried out with the 3D information. A hand region is extracted using the color information, and the operators gestures are recognized. 3D Finger pointing to select an appliance is measured using the cameras. Experiments verify the effectiveness of the proposed 3D measurement methods.

Construction of an Intelligent Room using Distributed Camera System

These days, computerization of our living environment is progressing, and the home appliances are becoming more intelligent and function-rich. On the other hand, the increase of their functions makes their operation complicated. For such appliances frequently used in everyday life, intuitive operation is desirable for users. Gestures, which we use frequently and intuitively in our everyday communication, can be one of such human-machine interfaces. Until now, many studies which recognize gestures from a sequence of images have been reported (V. I. Pavlovic et al., 1997). A shortage of many of the gesture recognition studies is that the place where gestures can be recognized is limited and thus they are not suitable for practical applications. There is a trend of studies to make a room or some space itself intelligent (A. Pentland, 2000), (M. Coen, 1999), (Brumitt, B. et al., 2000), (T. Mori & T. Sato, 1999), (J. H. Lee & H. Hashimoto, 2002). This approach is effective for realizing natural human-machine interface. Gesture recognition is one of the important technologies for such an intelligent room or space. We are also constructing an intelligent room in which an operator can control home appliances such as a TV set with intuitive gestures (K. Irie et al., 2004). Characteristics of our intelligent room are that, it focuses on operation of home appliances, it allows users to operate without restriction to the place and without having any instrument, and it uses pan-tilt-zoom cameras only as sensor information. In this paper, we discuss three dimensional (3D) measurement of the intelligent room using a distributed camera system, especially 3D measurement of hand waving and finger pointing.

Skin Color Registration Using Recognition of Waving Hands

This paper proposes skin color registration using the recognition of waving hands. In order to recognize hand gestures from images, skin colors are useful information. The proposed method can register skin colors simply and quickly because it uses just a few waves of the hand. The method consists of 2 steps. First, the regions of the waving hands are extracted from low-resolution images without using color information. Second, the color values of the extracted regions are classified into background colors and hand colors depending on time series of color images. The color information classified as hand colors is registered as skin colors. The proposed method is robust against lighting conditions and individual differences in skin color, because the skin color is registered as an adapted skin color in each case. Several experiments are conducted to demonstrate the effectiveness of the proposed method.

2A1-D11 Skin Color Registration for Gesture Recognition in an Intelligent Room

This paper proposes skin color registration in an intelligent room. We apply the gesture recognition to the operation of the household electrical appliances. In order to recognize the gestures, the skin color information is used in this system. Existing method we had constructed took time to register the skin color. That means, it was difficult to operate household electrical appliances smoothly. Therefore we developed the new method that improved the skin color registration, it recognize the gestures more smoothly by any skin colors. The new method uses a result of waving hands recognition that identifies an operator. It was confirmed that the method could resister the skin color smoothly and it was valid for an intelligent room system.

Subtraction Stereo - A stereo vision suitable for surveillance applications -

We demonstrate ldquosubtraction stereordquo, a simple and robust stereo vision algorithm that is suitable for surveillance applications such as detection of pedestrians.