Canny邊緣偵測立方積分內插法 之視訊格式轉換器設計

DESIGN OF A SCAN CONVERTER USING THE CUBIC CONVOLUTION INTERPOLATION WITH CANNY EDGE DETECTION

劉旭桓(Hsu-Huan Liu)

指導教授 ： 蔡明傑

大同大學/電機資訊學院/通訊工程研究所/碩士(2010年)

摘要

現今的數位顯示技術中，由於螢幕的尺寸越來越大，為使影像縮放可以盡量不失真，內插演算法便成為非常重要的技術。傳統的演算法有模糊或格狀效應的缺點。為了改善這些問題設計出很多邊緣偵測的方法，但大多數邊緣偵測設計架構較複雜，而且在影像縮放表現上也有所差異，因此為了要提高影像品質，本論文使用Canny邊緣偵測法來偵測出邊緣，並針對邊緣及非邊緣分別套入最鄰近內插法及立方積分內插法，以改善邊緣的銳力度及影像的清晰度。研究結果顯示本研究所提出的方法在影像的清晰度與邊緣銳利度上都優於傳統演算法，本實驗並使用Altera CycloneII FPGA來完成硬體的實現，將影像由SVGA（800×600）格式轉換至WXGA（1280×768）格式顯示在液晶面板上。

關鍵字

非適應內插演算法 ； 格狀效應 ； Canny邊緣偵測 ； 影像相似度 ； 影像銳利度

並列摘要

With the increasing use of digital monitor technique, requirements for image interpolation have become more critical. Especially, the development of LCD screen sizes has grown up. Conventional interpolation methods have several drawbacks, such as blurring or blocky effects. Many edge detect methods have been widely used to avoid these problems but are notorious for high complexity and cost. Therefore, in order to enhance quality of image. In the thesis, we used Canny edge detection to fined image edge, the part detected transforming by nearest neighbor interpolation the other part transforming by cubic convolution interpolation. This method improves sharp of edge and resolution of image. The FPGA implementation of the proposed video scaling algorithm is from SVGA(800×600) to WXGA(1280×768) on the TFT-LCD plane at the real time video.

並列關鍵字

Non-adaptive interpolation ； blocking effect ； Canny edge detection ； PSNR ； sharpness

參考文獻

[2] A. Pratt, and K. William, Digital image processing, Wiley, N.Y., 1978.

[4] B. Scarborough, and J. Blaine, Numerical mathematical analysis, Johns Hopkins Press, Baltimore, 1985.

[5] R. Keys, “Cubic Convolution Interpolation for Digital Image Processing,” IEEE Trans. Signal Processing, vol. 29, pp. 1153 - 1160, 1981.

[7] J. W. Hwang and H. S. Lee, “Adaptive Image Interpolation Based on Local Gradient Features,” IEEE Signal Processing Letters, vol. 11, pp. 359-362, 2004

[8] X. Li, and M. T. Orchard, “New edge-directed interpolation,” IEEE Trans. Image Processing, vol. 10, pp. 1521 – 1527, 2001.