Fast median filters using separable sorting networks

Noise removal is one of the major requirements in image, speech and signal processing applications. Impulse noise removal in image processing uses median filters. For edge preservation in image processing this acts as one of the best alternative non-linear technique to linear filtering. Real time hardware implementation of median filters has major concern of sorting networks. Efficient VLSI implementation of sorting network in terms of hardware complexity is of greater importance. This work provides a comparison of existing six data comparators and also proposes three modified data comparators in terms of their hardware complexity, area, power and speed. All the comparators were designed using verilog HDL and were targeted for xa6slx4-3-csg225using Xilinx ISE 9.2i FPGA design suite. From the results Modified Twos Complement Based Data Comparator is the minimum area required architecture with maximum combinational path delay and also with minimum number of LUTs used. The drawback of this architecture is the maximum memory requirement. The Modified Multiplexer Based Data Comparator and Modified Decoder Based Data Comparator architectures are suitable for memory efficient design.

Download Full-text

Model Predictive Torque Control for Multilevel Inverter fed Induction Machines Using Sorting Networks

IEEE Access ◽

10.1109/access.2021.3052129 ◽

2021 ◽

Vol 9 ◽

pp. 13800-13813

Author(s):

Kristof Bandy ◽

Peter Stumpf

Keyword(s):

Induction Machines ◽

Torque Control ◽

Multilevel Inverter ◽

Sorting Networks

Download Full-text

Feasibility of Total Variation Noise Reduction Algorithm According to Various MR-Based PET Images in a Simultaneous PET/MR System: A Phantom Study

Diagnostics ◽

10.3390/diagnostics11020319 ◽

2021 ◽

Vol 11 (2) ◽

pp. 319

Author(s):

Chan-Rok Park ◽

Seong-Hyeon Kang ◽

Young-Jin Lee

Keyword(s):

Image Quality ◽

Noise Reduction ◽

Total Variation ◽

Parallel Imaging ◽

Pulse Sequence ◽

Reduction Algorithm ◽

Median Filters ◽

Reduction Techniques ◽

Imaging Results ◽

Noise Reduction Algorithm

Recently, the total variation (TV) algorithm has been used for noise reduction distribution in degraded nuclear medicine images. To acquire positron emission tomography (PET) to correct the attenuation region in the PET/magnetic resonance (MR) system, the MR Dixon pulse sequence, which is based on controlled aliasing in parallel imaging, results from higher acceleration (CAIPI; MR-ACDixon-CAIPI) and generalized autocalibrating partially parallel acquisition (GRAPPA; MR-ACDixon-GRAPPA) algorithms are used. Therefore, this study aimed to evaluate the image performance of the TV noise reduction algorithm for PET/MR images using the Jaszczak phantom by injecting 18F radioisotopes with PET/MR, which is called mMR (Siemens, Germany), compared with conventional noise-reduction techniques such as Wiener and median filters. The contrast-to-noise (CNR) and coefficient of variation (COV) were used for quantitative analysis. Based on the results, PET images with the TV algorithm were improved by approximately 7.6% for CNR and decreased by approximately 20.0% for COV compared with conventional noise-reduction techniques. In particular, the image quality for the MR-ACDixon-CAIPI PET image was better than that of the MR-ACDixon-GRAPPA PET image. In conclusion, the TV noise-reduction algorithm is efficient for improving the PET image quality in PET/MR systems.

Download Full-text