A Theory for Multiresolution Signal Decomposition: The Wavelet Representation
IEEE Transactions on Pattern Analysis and Machine Intelligence
IEEE Transactions on Pattern Analysis and Machine Intelligence
Hi-index | 0.00 |
FPD(Flat panel displays) are becoming widely acceptable and popular as display devices growing at a rapid pace and making it a promising investment opportunity. Display quality is part of the final FPD inspection process prior to shipping. FPD devices have non-uniform background intensity, and shape and intensity of defects are very various. Due to the uneven illumination of panels it is difficult to detect the defects, Therefore this paper proposes a uneven illumination compensation method using wavelet based multi-resolution structure. This is first done by decomposing the image into multi-resolution levels and selecting some band frequency sub-bands by removing lower sub-bands and higher sub-bands corresponding to low varying signal and high frequency random noise, respectively, resulting in flat background images in reconstruction. A simple binary thresholding technique is then used to separate the defective regions from the restored image. Finally, blob analysis as post-processing is carried out to get rid of false defects.