視覺密碼學這個新興的密碼學領域，利用了人類視覺系統的特性直接對重疊的影像進行解密，它不須要任何密碼學的知識，也沒有複雜的計算過程。在安全性上，也可以確保截取者無法從這些個別的分享影像中察覺到任何與機密影像相關的紋理。自從Naor及Shamir在Eurocrypt'94論文中提出了基本的視覺密碼模型後，陸續有不少學者發表了相關的研究。然而這些研究大部份集中在黑白影像的討論上，少有針對灰階及彩色影像提出作法。本篇論文，我們將根據以往視覺密碼的研究，加上半色調技術及分色原理，提出幾種灰階影像和彩色影像的視覺密碼的作法。我們的作法不但延續了黑白視覺密碼直接利用視覺系統解密，無須大量運算的優點，對於先前有關黑白視覺密碼學的研究（如t out of n Threshold Scheme）亦可以輕易的套用在我們的方法上，而應用在灰階及彩色影像的製作上。 Visual Cryptography, which is an emerging cryptography technology, uses the characteristics of human vision to decrypt the overlapping images. It needs neither cryptography knowledge nor complex computation. In the regard to the safety, it ensures that hackers can't perceive any clues about the secret image on each cover image. Since Naor and Shamir proposed the basic model of visual cryptography in the Eurocrypt'94, several researchers published their related researches continually. Most of these studies, however, concentrated on black-and-white images; few of them proposed methods for processing gray-level and/or color images. In this paper, we will propose several methods for gray-level and color visualization cryptography based on the past studies of black-and-white visual cryptography and the halftone technology and color decomposition method. Our approaches do not only keep the advantage of black-and-white visual cryptography, which exploits the visual system to decrypt the secret image without huge computing, but also have the backward compatibility with the previous studies of black-and-white visual cryptography, such as t out of n threshold scheme can also be able to apply to gray-level and color images easily.