A Blur Image Blind Identify Algorithm Based on the Edge Feature

In view of the blur operation in the image tamper, a blur image blind identify algorithm based on the edge feature was proposed. Firstly, the algorithm made the logarithm operation to the image according to the edge feature of out-of-focus blur and artificial blur, and separated image´s the incident component and the reflection component. Then it used homomorphism filter to the wavelet region, and obtain the low frequency sub-graph and the high frequency sub-graph under different criterion using wavelet transformation´s multi-criteria decompose. Finally, the nature edge of image was corroded by mathematics morphology´s corrosion operation, and the blur tamper edge was obtained. The experimental result showed that the algorithm can identify effectively the image that was tampered with blur, and can locate the blur operation trace.