應用色彩空間之位移圖像於液晶光電辨識系統

Application of HSV Color Space for Optoeletronic Recognition System with Shifted Training Images

林紹達(Shao-Da Lin)

指導教授 ： 陳祖龍

元智大學/電機通訊學院/光電工程學系/碩士(2012年)

摘要

本論文中是以Minimum average correlation energy (MACE)配合影像位移找出最小平均相關能量的參考函數。本實驗先將彩色影像轉換HSV色彩空間的三個分量，此三個分量圖同時作旋轉(從5o到355o，每5o為間隔)和位移(x=-5到x=5 間隔為1個像素，y=-5到y=5 間隔為1個像素)。而後，將各自的三個分量原圖與處理後分量圖作MACE運算和線性相關，並找出最小旁辦能量所對應的參考函數並記錄其位置與相關數據，再進行分析。由此法我們使得相關峰值更尖銳及降低旁瓣能量，減少判斷錯誤，提升辨識效率。

關鍵字

影像辨識 ； 光電

並列摘要

This thesis is based on Minimum average correlation energy (MACE) with the image displacement to find the reference function of the minimum average correlation energy. First, we transform the color image into three HSV color space components. Then, three components are rotated (from 5o to 355o, the interval is 5o) and shifted (x = -5 to x = 5, the interval is 1 pixel, y =-5 to y = 5, the interval is 1 pixel) at the same time. Finally, we compute the original components and process components by using MACE method and linear correlation. Through the computation we will get the reference functions of the minimum sidelobe energy and record the relevant data. By the rotated and shifted processes, we could make the correlation peak more sharper and decrease the sidelobe energy, so as to increase the recognition ability and reduce false alarm rates .

並列關鍵字

Optoeletronic Recognition System

參考文獻

[1]F. T. S. Yu, Optical Information Processing, Wiley-Interscience, New York, 1983.

[2]A. VanderLugt, "Signal detection by complex spatial filtering." IEEE

T Inform Theory, vol. 10, 2, pp. 139-145, 1964.

[4]F. T. S. Yu and X. J. Lu, "A real-time programmable joint transform

correlator." Opt. Comm. 52, pp. 10-16, 1984.