Yusuke Yazu

On an image processing of eye blinking to monitor awakening levels of human beings

The main goal of the present study was to design and examine a system to monitor awakening levels of human beings using the image processing technique to recognize eye blinking. This problem was divided into two sub-problems; to obtain images by decreasing noise and simplify the images, and to detect whether eyes were closed or opened. A filtering technique and the automatic threshold determination method were used to solve the former sub-problem and the automatic detection system was designed for the latter sub-problem. To examine the reliability of the designed system, a comparison between the result detected by the system and that detected by a sensor using electroencephalograms (EEGs) was made.

Evaluating dynamic changes of driver's awakening level by grouped /spl alpha/ waves

Describes the system of automatically detecting grouped /spl alpha/ waves of EEGs using a convolution with special weighting factors such as moving average methods. A drop of human awakening level has a close relation to appearances of grouped /spl alpha/ waves. Then using this system, it is possible to detect drivers low awakening condition and prevent traffic accidents caused by this condition. The system has two important points. The first, because actual data contain various kinds of noise, this system separates grouped /spl alpha/ waves from those. The second, the system detects this condition as soon as grouped /spl alpha/ waves appear. As the result of analyzing actual EEGs taken during driving a car, we find out the fact that intervals of grouped /spl alpha/ waves appearance fluctuate.

Integrated environmental management process applying genetic algorithm

Abstract This study proposes a Total Environmental Management System using a Genetic Algorithm (TEMS-AGA), which takes into account the possible effects on the environment together with other management concerns such as quality, lead time, and cost. In TEMS-AGA, a problem is modeled by an AND–OR-type tree structure in its structural aspects and by a list of symbolic expressions in its quantitative aspects. TEMS-AGA evaluates the activity (policy, project) using a symbolic manipulation technique. GA is used to search a compromising better solution. Experiments showed that the determination of sub-optimal combinations of activities for a problem with multiple facet can be made by GA.

Dynamic changes of human systems under a simple task of HCI

Conclusions From the analyses and considerations, the conclusions of this paper are followings; u (1) We measured I/O responses of human system, states of human systems like amounts of grouped a waves, cardiac rhythms and winking under a simple HCI task. (2) The all measured data are changing dynamically. (3) It is possible to represent the response of human systems as a function of factors indicating states of human systems. (4) Amounts of grouped a waves, cardiac rhythms, and numbers of winking are one of the good factors to determine the states of human systems. (5) A results of (4) show the importance of studies about human states to design a suitable method to future HCI. (6) A flexibility or fluctuation is one of important characteristics of future HCI.

Possibility of ECGs to improve reliability of detection system of inclining sleep stages by grouped a waves

In order to prevent traffic accidents, developing a system which find out sleepy states of drivers and warning them the dangerous state. W e have found that special EEG8 called grouped a waves appear in sleepy state of drivers, and defined the sleepy states characterized by grouped a waves as ’inclining sleep (low awake) stage‘, And then, we made a subsystem to find out inclining Sleep stages of drivers using grouped a waves. The first kind of error about the subsystem is 2% and tne second kind Of that is 25”35%. In order to improve the reliability of the system, we analyze ECG6 as detail as EEGs during driving. As the results, EEGs is more suitable to sleepy states of drivers and ECGs are also used t o improve the reliability of such a detection-subsystem. 1 . INTRODUCTION It is important to develop a system to prevent traffic accidents caused by drivers‘ dozing off. We have analyzed electroencephalograms (EEGa) during driving a car or other simple works. As the results. we find out the fact that a special EEC8 called as grouped a waves appear in most cases. Then we defined the state which is characterized by the appearance o f grouped a waves as ‘inclining sleep stage’ and made a system which automatically detects inclining sleep stage using grouped a waves‘’. The ratio which the system can not detect inclining sleep is 2% and the ratfo which the system detects non-inclining sleep stage as inclining sleep stage is about 25%-35%. Since it is required not to overlook sleepy stages to such a detection system, the system w e made satisfied such property. But we wish to refine the system more reliable, and then we measure not only EEGs but also electrocardiographs (ECCs) and analyze them. Because i t is reported that ECGs are deeply concerned with fatigue of human beings‘’. in t h e f i led of ergonomics. In this paper, we analyze both of EEGs and ECGs, observe the relations between them, and discuss the possibility of ECGs to refine the detection system of fnclining sleep stages.

A study on shifting time to low awakening conditions on monotonous VDT works

Publisher Summary This chapter presents the study of three kinds of monotonous video display terminal (VDT) work that have different interface. During repeated monotonous work, which needs less physical exercise, most of workers shift to low awakening condition. Then the characteristic of shifting time to this condition are recorded. Experiment 3 this work is an analogous behavior which is more continual and tends to shift to low awakening level. Experiment 1 this work is a digital behavior, its rhythmical and need a little judgment. However, they get used to this work then shift to low awakening level. In experiment 2 they need ability of calculate which is a high-level brain activity. They feel not only monotonous but also fatigue and shift to low awakening level. In term of the appearance time of grouped a waves there are few difference among three types of interaction. During monotonous work which receives less stimulation from outside and needs less physical exercise we confirm that electroencephalography that reflect physical condition directly change from 20 to 30 minutes later and awakening level go down objectly. They cannot prevent drowsiness. There are two reasons why we shift to low awakening level. They are monotonousness and fatigue.

Character and application of winking pattern as alternative index for grouped /spl alpha/-wave

Investigating drowsy states with brain waves has been standard until now. The authors considered winking patterns instead of brain waves, and here they investigate the character of winking patterns and discuss the relation between sleeping state and winking patterns. As a result, it was confirmed that the difference of duration of eyelid movement increases gradually, and it is necessary to give mental tasks to the subjects for car driving simulation. Moreover, from these results the authors conclude that the winking pattern may be used as alternative index for grouped /spl alpha/-waves.<<ETX>>