ROI Based Error Concealment of Object Based Image Using Data Hiding in JPEG 2000 Coding Pipeline

The transmission of multimedia signals, such as image over noisy wireless channels may cause inevitable errors that might severely degrade the visual message. Error concealment (EC) techniques by exploiting inherent redundancy reduce visual artifacts through post processing at the decoder. In this paper, we present a data hiding scheme for image error concealment in JPEG 2000 coding pipeline using integer wavelet, quantization index modulation (QIM) and region-of-interest (ROI) coding functionality. To restore the human interested portion of image, the information (image digest) of ROI is embedded into the region-of-background (ROB) using data hiding. Threshold based image segmentation together with morphological operations are used for finding ROI, while halftoning technique is used to obtain image digest from ROI. The simulation results show that the scheme results in efficient performance measure with various lost blocks in ROI.