赫夫轉換類神經網路於震測圖型之分析

Abstract

在本論文中提出一種可用於赫夫轉換的類神經網路架構，特別是用於震測訊號的分析當中，針對直接波（直線）與反射波（雙曲線）的辨認而設計。 對於震測訊號的分析與處理，直接波（直線）與反射波（雙曲線）的辨認至為重要，因此本文所提之方法有助於震測訊號的分析，以利後續之處理。

A neural network model for Hough transform is proposed. This model can be applied to seismic signal processing and recognition of directed wave (line) and reflected wave (hyperbola) on seismic signal. For seismic signal analysis and processing, the detection of directed wave (line) and reflected wave (hyperbola) are very important. Our method provides convenience to the seismic signal processing.