TY - GEN
T1 - FPGA implementation of AIQ based Daubechies filter Bank for Medical Image Retrieval
AU - Samantaray, Aswini Kumar
AU - Sahoo, Satyajit
AU - Rahulkar, Amol D.
N1 - Publisher Copyright:
© 2024 IEEE.
PY - 2024
Y1 - 2024
N2 - Wavelet transform techniques are crucial to image processing, particularly in medical imaging, where the need for precise feature extraction is critical. This article describes the design and FPGA implementation of an Algebraic Integer Quantization (AIQ) based Daubechies-6 (DAUB6) wavelet filter bank, required for efficient medical image retrieval. The proposed AIQ-based orthogonal wavelet filter bank helps to increase the retrieval accuracy by optimizing the frequency and spatial resolution, important for medical images. The current filter bank design is executed on a Field Programmable Gate Array (FPGA) using Xilinx VirtexE FPGA board to leverage its parallel processing capabilities, ensuring low power consumption. The FPGA implementation is validated using a dataset of medical images such as NEMA, OASIS and EXACT09. Through extensive simulations and hardware synthesis, the proposed system demonstrates significant improvements in retrieval efficiency and accuracy compared to existing wavelet filter banks.
AB - Wavelet transform techniques are crucial to image processing, particularly in medical imaging, where the need for precise feature extraction is critical. This article describes the design and FPGA implementation of an Algebraic Integer Quantization (AIQ) based Daubechies-6 (DAUB6) wavelet filter bank, required for efficient medical image retrieval. The proposed AIQ-based orthogonal wavelet filter bank helps to increase the retrieval accuracy by optimizing the frequency and spatial resolution, important for medical images. The current filter bank design is executed on a Field Programmable Gate Array (FPGA) using Xilinx VirtexE FPGA board to leverage its parallel processing capabilities, ensuring low power consumption. The FPGA implementation is validated using a dataset of medical images such as NEMA, OASIS and EXACT09. Through extensive simulations and hardware synthesis, the proposed system demonstrates significant improvements in retrieval efficiency and accuracy compared to existing wavelet filter banks.
UR - https://www.scopus.com/pages/publications/86000203196
UR - https://www.scopus.com/inward/citedby.url?scp=86000203196&partnerID=8YFLogxK
U2 - 10.1109/ASPCC62191.2024.10880936
DO - 10.1109/ASPCC62191.2024.10880936
M3 - Conference contribution
AN - SCOPUS:86000203196
T3 - 2024 IEEE 1st International Conference on Advances in Signal Processing, Power, Communication, and Computing, ASPCC 2024
SP - 38
EP - 42
BT - 2024 IEEE 1st International Conference on Advances in Signal Processing, Power, Communication, and Computing, ASPCC 2024
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 1st IEEE International Conference on Advances in Signal Processing, Power, Communication, and Computing, ASPCC 2024
Y2 - 19 December 2024 through 21 December 2024
ER -