Using JPEG2000 ROI Progressive Coding for Mammograms Compression

• Main Authors: Jung-Ping Li, 李榕苹
• Other Authors: Pau-Choo Chung
• Format: Others
• Language: en_US
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/32056760113214973179

碩士 === 國立成功大學 === 電機工程學系碩博士班 === 92 === 　　The mammography is one of the most effective screen methods in detecting the tiny calcifications. Due to that the size of each mammogram is very large and different quality level of mammogram image obtained with different compression ratios are adopted for different diagnosis requirements in current hospital. Therefore, duplicated storage required and maintenance difficulty was happened. In this thesis, the ROI progressive coding in JPEG2000 is used for compressing to solve the foregoing problem in the hospital and use the region of interest (ROI) coding to make sure the display with high quality in lesions. The system analyzes the relationship between the number of compressed bitplanes in non-lesion and the quality of image. Thus, the system programmer can design the number of compressed bitplanes in each layer depending on referring to the re-diagnosis by the doctor and the formula designed in the thesis.