Shin Yamamoto

Detecting method for drivers' drowsiness applicable to individual features

It is inevitable for driver assist and warning systems to consider the drivers state of consciousness. Drowsiness is one of the important factors in estimating the drivers state of consciousness. A method to extract the drivers initial stage of drowsiness was developed by means of the blink measurement irrelevant to the surrounding brightness and individual characteristics with motion pictures processing. The result was that an increase of the long eyelid closure time was the key factor in estimating the initial stage of drivers drowsiness while driving. And the state of drowsiness could be presumed by checking the frequencies of long eyelid closure time per unit period.

Wide dynamic range vision sensor for vehicles

The dynamic range of brightness on road scenes is very wide, because the lighting condition dynamically varies with various weather and road conditions. Therefore, the dynamic range of conventional TV cameras is insufficient to input the images of road scenes. The authors have developed a method for expanding the dynamic range of TV cameras. Also, they have developed an experimental vision sensor system with a wide dynamic range based on the method applicable to the vision systems for vehicles. The effectiveness of the sensor in comparison with conventional TV cameras was confirmed from the experiments on highways under various lighting conditions.<<ETX>>

Method of detecting drowsiness level by utilizing blinking duration

The purpose of this study is to detect the drowsiness of drivers to prevent traffic accidents. We have developed a system which measures the blinking duration of a driver by processing his facial image and estimates his drowsiness level. This paper presents an imaging unit which is less affected by ambient light, a method of measuring the drivers blinking duration which is independent of his facial features, and a method of estimating his drowsiness level from a change in his blinking duration.

Real-Time Range Measurement Device for Three-Dimensional Object Recognition

This correspondence describes a real-time range measurement device for detecting the range, shape, and position of a three-dimensional object by a light stripe projection method. The device consists of a light stripe projection unit, a light stripe scanning control unit, a TV camera, a signal processing unit, and a microcomputer. While a laser light stripe is projected and scanned on an object at a high speed, range measurements and their calculations are done with special hardware. Range data are obtained by triangulation. This device can detect the range and shape of the object placed in a 60 cm square area at a distance of 100 cm in 490 ms by 48 × 50 points. The measurement accuracy of the device is ±2.0 cm. A sorting system for machine parts using this device has also been developed. By use of a unique range gating method, the system can discriminate three types of machine parts on a conveyor belt in 4.0 s.

Electrical environmental characteristics for automotive electric systems

The electrical environment is one of the most important factors in the design of high reliability electronic systems mounted on automobiles. Studies were made of the generation mechanisms and characteristics of both transients and high-frequency conducted noise generated in automotive electrical components. The high-frequency conducted noise data were obtained in terms of amplitude-frequency through real-time statistical treatment of the acquired data. The most important transient for automotive electronic systems has been found to be the alternator load dump transient caused by battery disconnection, and the most influential high-frequency conducted noise is caused by contact breaking. The high-frequency conducted noise characteristics can be expressed in the amplitude-frequency relation in the frequency range of 100 kHz-80 MHz.

Effectiveness of video camera dynamic range expansion for lane mark detection

Insufficiency of dynamic range of conventional video cameras is an important problem in realizing a robust vision system for autonomous vehicles, because road scenes have very wide dynamic range of brightness. To solve the problem, we have developed an experimental visual sensor having an expanded dynamic range, which is approximately 30 times as wide as that of a conventional CCD camera. By applying the developed sensor to lane mark detection, the effectiveness of the dynamic range expansion in comparison with a conventional CCD camera was evaluated. The successful detection rate with the developed sensor was 6% greater than that with the conventional camera on an average in daytime, and the effectiveness of the dynamic range expansion was confirmed.

RF Conducted noise measurements of automotive electrical and electronic devices using artificial network

Two RF conducted noise measurement systems have been developed for laboratory measurements of automotive electrical and electronic devices. One is a noise emission measurement system for measuring noise level from the electrical devices while the other is a susceptibility measurement system for measuring the susceptibility level of the electronic devices to the noise. The formers function is to measure a histogram of noise level, and the latter functions both to measure the malfunction level and to judge a malfunction mode of the electronic devices. These measurement systems are applicable to the frequency of 150 kHz-60 MHz. These are for laboratory measurements performed with the measurement systems connected to an artificial network of the automotive electrical network. This artificial network represents electrical networks used on Japanese compact passenger vehicles.

Properties of High-Frequency Conducted Noise from Automotive Electrical Accessories

This paper describes generation mechanisms and electrical characteristics of high-frequency conducted noise from electrical contacts and accessory drive/control motors, which may cause logic errors in automotive electronic devices. High-frequency conducted noise due to contacts in both horn sounding and lighting circuits results from spark discharge initiated by breakdown across the contacts; and a showering arc which sustains spark discharge for hundreds of microseconds was observed at contacts in lighting circuits. High-frequency conducted noise due to accessory drive/control motors was caused by reactance voltage fluctuations and commutator sparks. To analyze electrical characteristics of high-frequency conducted noise, a measuring system for amplitude distributions in both the time domain and the frequency domain was developed, and an attempt was made to fit a distribution pattern to observed data in each domain. In the time domain, the noise voltages due to contact breaking were distributed between 360 and 530 V. The distribution pattern in the time domain fitted closely to the normal distribution curve, while the pattern in the frequency domain fitted to the lognormal distribution curve.

Robustness of lane mark detection with wide dynamic range vision sensor

This paper presents the improvement of robustness of the vision system for lane mark detection using the wide dynamic range vision sensor. The performance of the lane mark detection was evaluated by experiments for the road scenes such as a tunnel exit or bridge crossing road, in comparison with the vision system using a conventional TV camera. A conventional edge-based image recognition technique was used for lane mark detection. The experimental results have confirmed the robustness of the vision system with wide dynamic range vision sensor.

A new recognition method for stamped and painted alphanumerals

In the automobile production system, the recognition technique for the stamped character stamped on the painted surface is indispensable in part labelling and production control. When a stamped character is painted over, it often happens that undulations and humps are produced on the character contour. This paper discusses a recognition method for such stamped characters. In order to make the method robust toward variations in the character contour, the contour shape data is constructed viewing the character from the circumscribed rectangle of the character. Then, the contour features inherent to the character are extracted after smoothing, to perform the recognition. n nThe method was applied to stamped and painted 1000 samples of ten kinds of alpha-numerals and the recognition rate was examined. It is seen that the proposed method exhibits a higher recognition rate for all character types than the multiple similarity method. According to the experiment in which the experimental system based on the proposed method was applied to 3000 samples of 15 kinds of alphanumerals, the correct recognition rate was 99.8 percent. Although there were seven samples which were rejected, there were no misclassifications. Thus, the usefulness of the method was verified.