Energy Interpolated Mapping for Image Compression with Hierarchical Coding

Objectives: To achieve high compression rate by reducing the more redundant information in medical images. More compression rate results in less resource utilization and also reduces the processing overhead and time consumption. Methods/Statistical Analysis: In this approach, initially the medical image was decomposed through multiwavelet transform. Then a band selection procedure is performed on the obtained sub bands to select the bands which are noncorrelated. Thus, the redundant information existing in the bands will be reduced. Then, the selected bands were processed for energy based interpolation to select the features which are more informative and also to reduce the redundant information further. Next, the hierarchical coding was applied over the obtained features. Findings: Simulation results are carried out over various medical images and for every image, the quality was checked through PSNR and the performance was checked through processing overhead and computation time (sec). Compared with earlier approaches, the processing overhead of proposed approach observed to be less and the computation time also. Similar, the PSNR is observed to be high and MSE as low. Applications/Improvements: The proposed medical image coding system will be used in telemedicine applications where there is a need of efficient resource utilization to transmit the data with fewer resources.