Xiaowei Fu

An image despeckling approach using quantum-inspired statistics in dual-tree complex wavelet domain

Abstract This paper presents a novel despeckling algorithm to enhance image quality of medical ultrasound images. The proposed approach exploits coefficients in the dual-tree complex wavelet transform (DTCWT) domain to develop a new quantum-inspired thresholding function. The inter-scale correlation among the coefficients of different subbands and the intra-scale variance between the coefficient and its neighborhood in the same subband are utilized to develop a new thresholding function, which is further incorporated into a Bayesian framework to perform adaptive image despeckling. Experiments are conducted using both artificially generated and real-world medical images to demonstrate that the proposed approach outperforms the conventional image despeckling approaches.

Blind noisy image quality assessment using block homogeneity

Blind noisy image quality assessment aims to evaluate the quality of the degraded noisy image without the need for the ground truth image. To tackle this challenge, this paper proposes an image quality assessment approach using block homogeneity. First, motivated by the local smoothness characteristic of the image, a block-based homogeneity measure is proposed to estimate the statistics (e.g., variance) of the noise incurred in the image, based on adaptively selected homogeneous image regions. Second, an image quality assessment approach is proposed by exploitie above-mentioned estimated noise variance, along with the visual masking effect of the human visual system. Experimental results are provided to demonstrate that the proposed image noise estimation approach yields superior accuracy and stability performance to that of conventional approaches, and the proposed image quality assessment approach achieves consistent performance to that of human subjective evaluation.

Salient object detection from distinctive features in low contrast images

Saliency computational model with active environment perception can be useful for many applications including image segmentation, image compression, image retrieval, and etc. Conventional saliency computational models rely on handcrafted low level features, such as color or contrast. These models face great difficulties in low lighting scenarios, due to the lack of well-defined feature to interpret saliency information in low contrast images. In this paper, a new approach is proposed to detect salient object from low contrast images. The proposed approach explores the most distinguishable salient information in low contrast images based on low level features. Extensive experiments have been conducted to evaluate the performance of the proposed method against the state-of-the-art saliency computational models.

Robust passive autofocus system for mobile phone camera applications

Display Omitted We propose a robust passive autofocus system for mobile phone camera applications.We propose a novel modified Entropy based contrast measurement to determine the degree of image sharpness.We propose an effective peak search strategy to test the performance. A robust autofocus system is a ubiquitous function in todays mobile phone camera applications. However, due to the power consumption and size requirements, it is difficult for the autofocus function to be implemented into the design of mobile phone cameras. This paper presents a passive autofocus system with low computational complexity. This system uses a novel contrast measurement to determine degree of image sharpness, which can better reflect the information about image discontinuities. In order to gauge the performance of this measurement, a modified peak search strategy is used in the experiments. The experimental results from several typical image sequences validate the effectiveness of the proposed method.

DTCWT based medical ultrasound images despeckling using LS parameter optimization

This paper presents a novel despeckling algorithm that can be used to enhance image quality in medical ultrasound images. Firstly, the log-transformed images are transformed by dual-tree complex wavelet transform (DTCWT). And then, we use a non-Gaussian statistical model with an adaptive smoothing parameter for ideal image signal in the transformed domain. According to Bayesian theory, the MAP estimator is obtained with a proposed adaptive threshold which has better despeckling performance by exploiting the interscale properties of wavelet coefficients. The proposed approach results in significant speckle reduction and preserve details of ultrasound images at the same time while the introduced distortions are not noticeable.

Chinese Characters Recognition from Screen-Rendered Images Using Inception Deep Learning Architecture.

Text recognition from images can substantially facilitate a wide range of applications. However, screen-rendered images pose great challenges to current methods due to its low resolution and low signal to noise ratio properties. This paper proposed a Chinese characters recognition model using inception module based convolutional neural networks. Chinese characters were firstly extracted using vertical projection and error correction; then it can be recognized via inception module based convolutional neural networks. The proposed model can effectively segment Chinese characters from screen-rendered images, and significantly reduce the training time. Extensive experiments have been conducted on a number of screen-rendered images to evaluate the performance of the proposed model against state-of-the-art models.

Fuzzy Approach Based on Quantum-Inspired MRF for SOFC Anode Optical Microscope Images Segmentation

Microstructural information acquired from image analysis can be used in cell modeling. In order to obtain more precise Solid Oxide Fuel Cell (SOFC) microstructure parameters, an adaptive fuzzy approach is developed for three-phase identification of YSZ/Ni anode Optical Microscopic (OM) images. A new quantum-inspired clique potential Markov random field (MRF) function is proposed to considerate spatial information in fuzzy logic model, where the space distance based weight is introduced to reflect the influence of neighborhood pixels. Simulated images and real SOFC anode OM images are used to compare the effectiveness and practicability of the proposed algorithm with others. Experiment results demonstrate that the proposed methods can accurately separate there-phase of SOFC OM images, which lays the foundation of subsequent microstructural parameters extracting.

A Quantum-Inspired Fuzzy Clustering for Solid Oxide Fuel Cell Anode Optical Microscope Images Segmentation

For better three-phase identification of Solid Oxide Fuel Cell (SOFC) microstructure, this paper presents a novel quantum-inspired clustering method for YSZ/Ni anode Optical Microscopic (OM) images. Motivated by Quantum Signal Processing (QSP), a quantum-inspired adaptive fuzziness factor is introduced to adaptively estimate the parameters of the spatial constraint term in the fuzzy clustering based on Markov Random Filed (MRF). Experimental results show that the proposed method is effective to identify the three phases. The combination of image processing and micro-investigation provides an innovative way to analyze the performance of SOFC.

GMM- based image segmentation approach for SOFC microstructure characterization

In order to accurately evaluate the microstructure parameters of Solid Oxide Fuel Cell (SOFC) electrode, this paper presents a novel image segmentation method based on Gaussian Mixture Model (GMM) to identify the three phases of electrode optical microscope image. Firstly, the spatial neighbor information is introduced into EM optimization algorithm to constrain the weighted probability distribution of each pixel. Secondly, for uncertain points whose probabilities of two components are close, the probability distribution of them is adjusted according to quantum-inspired adaptive weight. The experimental results show that the proposed method is effective to separate the three phases of electrode, and provide reliable data support for SOFC 3D reconstruction.

Coarseness-entropy based Gaussian Mixture Model for SOFC image segmentation

For the three-phase identification of Solid Oxide Fuel Cell (SOFC) electrode, this paper presents a novel segmentation method based on Gaussian Mixture Model (GMM). A coarseness-entropy adaptive factor is defined to incorporate the spatial information based on Markov Random Filed (MRF) into GMM. Furthermore the proposed method defines can control the trade-off between robustness to noise and effectiveness of preserving the details. Experimental results show that the proposed method outperforms the compared method on three-phase microstructure identification.