<![CDATA[Image attributes, such as shape, orientation, size, texture, gradient, luminance, brightness, andcontrast can be easily recognized by human visual system. All these attributes contribute to thecharacteristics of an image or image features, which reveal uniqueness of the image. Once theimages get corrupted due to compression or transmission, the unexpected artificial features, whichare not existing in the origi"al images are introduced. The characteristics of these unexpectedfeatures known as artifact characteristics vary from one artifact to another By knowing the originalimage features, un-expected features can be detected. There are a number of feature extractiontechniques that can be used to extract image features, in order to observe the presence of artifacts.These techniques can be generally classified into two categories, as spatial domain based andtransform domain based. This paper presents a review of the existing spatial domain artifactextraction techniques using no-reference (withoul comparison to original image) as they arecomputationally inexpensive and suitable for in-service quality monitoring.Keywords: Artifact, feature extraction, error, blocking, blur, ringing, masking, lost blocks.]]>
Copyrights © 2007
