| Because of its characteristics of good resolution, high working efficiency, easy operationand no damage to detected things, Geological radar is widely used in different kinds ofengineering field, especially more frequently in the tunnel engineering application .The mainpurpose of this article is to study the geological radar application technology and imageprocessing algorithm, as well as the in-depth analysis and research, including noise reductionand edge detection. Combining the two kinds of processing technology, the engineeringtechnical personnel would explain the geological radar image more accurately and effectively.Geological radar is different from sounding radar.The application environment andapplication object of the Geological radar are relatively complicated. Therefore, the hardwarestructure is carefully studied in this paper firstly. The working principle of geological radar isalso deeply analysised. Then, the geological radar detection method as well as the geologicalradar image characteristics of typical geologic body are studied in order to well understandthe application of geological radar. On this basis, some detection examples of geological radarin engineering application is lead in to intuitively and concretely illustrate the geological radarapplication technology.The geological radar application environment is relatively complicated. So the collectedgeological radar image unavoidably includes noise interference, and the main informationfeature of geological radar image is the lineups of target layer. Therefore, the interferencefactors of geological radar image is firstly analysised in this paper, after that denoising andedge detection is conducted. Denoising mainly eliminates the influence of noise to the imageresolution, while edge detection mainly makes the edge of the lineups more clear. In view ofthe geological radar image feature, wavelet transform has a good time-frequency localizationcharacteristics, and always denoises effectively. But it fails to keep lineups of the geologicalradar images goodly; While K-L transform improves the same phase axis in the images, itfails to take count of image frequency domain signal-to-noise ratio. So on the base ofimproving traditional wavelet transform and K-L transform, the wavelet domain slant K-Ltransformation is put forward to denoise the geological radar image. After that, multi-scaleand multi-element detection mathematical morphology edge detection is conduct to denoiseimage, which can effectively depict the lineups of the geological radar images. The image processing algorithms put forward in this paper has a lot of use value in engineeringexplain .By combining with wavelet domain slant K-L transform results and multi-scale andmulti-element detection mathematical morphology edge detection, the engineering technicalpersonnel would explain the object being detected more accurately and effectively andachieve a better results. |
PDF Full Text Request