[in Japanese]

View full abstract

Traffic signs have characteristics color, spatial arrangement, and texture. In this paper we describe how to select the video frame in which the traffic sign has the largest size. We select it using three methods color filtering, a detection window, and textural analysis. Experimental results show that our combination of different methods works well.

View full abstract

This paper presents novel methods of recognizing vehicle license plates robustly. To acquire fine images of quickly passing vehicles under a wide range of illumination conditions, we developed a sensing system having superb characteristics. We expanded the dynamic range and eliminated the blurring of images of fast moving vehicles by synthesizing a pair of synchronized images with different intensities. Furthermore, to extend the flexibility of the positioning of the TV camera, we propose a recognition algorithm which can be applied to inclined plates. The performance of the integrated system was investigated on real images of highly inclined plates of vehicles captured under various illumination conditions. The recognition rate of 97% shows that the developed system is effective for license plate recognition.

View full abstract

A software solution for GPS baseband processing using a 32-bit embedded microprocessor is described. Since spectrum de-spreading takes up most of the data processing time in the GPS baseband processing system, we propose specifications, algorithms, and implementation methodologies to reduce the load. Our software solution for GPS baseband processing was verified by creating an experimental model equipped with a 32-bit embedded microprocessor.This successfully received signals from 8 satellites simultaneously, and GPS baseband processing was performed by a CPU with a processing capacity of approximately 40 MIPS.

View full abstract

[in Japanese]

View full abstract

We are studying an intelligent robot camera that can shoot automatically as well as a human broadcast cameraman can. To elucidate a cameraman's zooming techniques, we recently examined our measurement methods and conducted a basic experiment while shooting still subjects. The results showed that our measurement methods were effective in faithfully gauging zooming characteristics. Furthermore, analyses of zooming actions revealed that, in the case of repeated shooting of the same scene, the velocity of the operating rod always followed almost the same shaped zooming velocity curve, and that the rates of size change of the subject for normalized zooming times were 0.62-0.91 (second half) in the zooming-in operation and 0.17-0.44 (first half) in the zooming-out operation.

View full abstract

[in Japanese]

View full abstract