| Project/Area Number |
08455177
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
情報通信工学
|
|---|
| Research Institution | KYOTO UNIVERSITY |
|---|
Principal Investigator |
SATO Toru Kyoto University, Graduate School of Informatics, Professor, 情報学研究科, 教授 (60162450)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KASAHARA Yoshiya Kyoto University, Graduate School of Informatics, Research Associate, 情報学研究科, 助手 (50243051)
KIMURA Iwane Osaka Institute of Technology, Faculty of Information Science, Professor, 情報科学部, 教授 (00025884)
|
|---|
| Project Period (FY) |
1996 – 1998
|
| Project Status |
Completed (Fiscal Year 1998)
|
| Budget Amount *help |
¥6,700,000 (Direct Cost: ¥6,700,000)
Fiscal Year 1998: ¥1,300,000 (Direct Cost: ¥1,300,000)
Fiscal Year 1997: ¥2,600,000 (Direct Cost: ¥2,600,000)
Fiscal Year 1996: ¥2,800,000 (Direct Cost: ¥2,800,000)
|
|---|
| Keywords | subsurface remote sensing / high-resolution radar / radar signal processing / discrete model fitting / inhomogeneous media |
|---|
| Research Abstract |
In subsurface radar applications, it is often experienced that the requirements for the resolution and the penetration depth contradict each other. In such a case, the compromise will be the use of around 500MHz, for which the target is on the order of a radar wavelength. It is thus hard to identify the shape of the object with conventional synthetic aperture method.
Our objective is to develop a robust high-resolution imaging algorithm applicable to targets whose dimension is close to the radar wavelength. In order to avoid instabilities in super-resolution techniques, we limit the number of freedom in the estimation by properly modeling the target as well as the parameters of the medium. Our algorithm currently deals with two- dimensional imaging based on a one-dimensional scan of the radar sensor, although both monostatic and multistatic cases can be treated.
At the first step, outstanding targets are estimated as a group of point targets. Discontinuity in the medium are also estimate
… More
d based on the consensus analysis of the received time series at each point during the scan. Non-linear least squares fitting is used to improve the model in an iterative manner in estimating the layer boundaries and the point targets. In the fitting, observed echo time series is compared with the estimated ones generated from the model by a ray tracing algorithm.
For each prominent target found in the first step, the shape estimation procedure is applied. The shape of the target is expressed in terms of the points which represents its outer boundary. These points are selected so that the distance between adjacent points are on the order of 1/10 of the radar wavelength, and they are connected smoothly with Lagrangian interpolation. The reflections from these connected parts are computed by the ray tracing method, and the edge refraction component is appended with the aid of physical optics. The non-linear fitting is again used to determine the target shape.
The major restriction of this imaging technique is that proper initial guesses are required for the model parameters in each non-linear fitting process. Automated procedures for finding such initial parameters have been devised, We have examined the performance of this imaging technique with numerical simulations and test site experiments. Less
|
