查詢結果分析
來源資料
頁籤選單縮合
題名 | A Research on Designing an Adaptive Image Edge Detector=適應性影像邊緣偵測器設計之研究 |
---|---|
作者 | 劉建宏; | 書刊名 | 中華工商專校學報 |
卷期 | 12 1995.12[民84.12] |
頁次 | 頁79-87 |
分類號 | 312.1 |
關鍵詞 | 影像邊緣偵測器; |
語文 | 英文(English) |
中文摘要 | 本文提出一種新的適應性邊緣偵測器來作退化或含雜訊影像之邊緣偵測。首先假 設因果自我回歸之影像模型,利用統計估測法則設計遞迴式濾波器,除了能使影像依序產生 最佳罩幕係數作影像還原外,並能去模糊及濾雜訊。再利用一階及二階偏導數法作尖銳化, 萃取邊緣特徵。 實驗中,我們處理實際影像與電腦合成影像,發現對 SNR 小之圖像具有去 模糊之效果,對 SNR 大之圖像則處理後更加清晰, 最後並與拉普拉斯零交越法作比較,發 現對複雜圖像之處理,在雜訊去除與影像品質上,皆優於後者。 |
英文摘要 | We present a new adaptive edgs detector to extract the edge from noisy or degraded image. First the causal AR image modle and the statistical mothod are employed to design the recursive filter, which not only generate the optimal mask weightings for image restoration but also reduce noise and blur, Furthermore, we employ a directionl derivative approach to sharp the image for edge extraction. In the experiment, we process the real and synthetic image. We find that the blur of the low SNR image is reduce and the high SNR image become high quality image. We also compare with Laplacian Gaussian Zero-crossing method and find our result is better than the latter in quality and noise reduction. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。