Browsing by Author "Gao, Huijun"

Now showing 1 - 15 of 15
  • Thumbnail Image
    Item
    Adaptive Charbonnier superresolution method with robust edge preservation capabilities
    (Journal of Electronic Imaging, 2013-12-16) Maiseli, Baraka J.; Liu, Qiang; Elisha, Ogada Achieng; Gao, Huijun
    Superresolution (SR) is known to be an ill-posed inverse problem, which may be solved using some regularization techniques. We have proposed an adaptive regularization method, based on a Charbonnier nonlinear diffusion model to solve an SR problem. The proposed model is flexible because of its automatic capability to reap the strengths of either linear isotropic diffusion, Charbonnier model, or semi-Charbonnier model, depending on the local features of the image. On the contrary, the models proposed from other research works are fixed and hence less feature dependent. This makes such models insensitive to local structures of the images, thereby producing poor reconstruction results. Empirical results obtained from experiments, and presented here, show that the proposed method produces superresolved images which are more natural and contain well-preserved and clearly distinguishable image structures, such as edges. In comparison with other methods, the proposed method demonstrates higher performance in terms of the quality of images it generates.
  • Thumbnail Image
    Item
    Adaptive Charbonnier Superresolution Method with Robust Edge Preservation Capabilities
    (International Society for Optics and Photonics, 2013) Maiseli, Baraka J.; Liu, Qiang; Elisha, Ogada Achieng; Gao, Huijun
    Superresolution (SR) is known to be an ill-posed inverse problem, which may be solved using some regularization techniques. We have proposed an adaptive regularization method, based on a Charbonnier nonlinear diffusion model to solve an SR problem. The proposed model is flexible because of its automatic capability to reap the strengths of either linear isotropic diffusion, Charbonnier model, or semi-Charbonnier model, depending on the local features of the image. On the contrary, the models proposed from other research works are fixed and hence less feature dependent. This makes such models insensitive to local structures of the images, thereby producing poor reconstruction results. Empirical results obtained from experiments, and presented here, show that the proposed method produces superresolved images which are more natural and contain well-preserved and clearly distinguishable image structures, such as edges. In comparison with other methods, the proposed method demonstrates higher performance in terms of the quality of images it generates
  • Thumbnail Image
    Item
    An automatic and cost-effective parasitemia identification framework for low-end microscopy imaging devices
    (IEEE, 2015-09-03) Maiseli, Baraka J.; Mei, Jiangyuan; Gao, Huijun; Yin, Shen
    In the detection of Malarial parasites from a patient, it is usually necessary to carefully examine the corresponding blood-slide smear and distinguish the infected and healthy Red Blood Cells (RBCs). If this process is done manually, as evidenced in common traditional approaches, the following challenges may be encountered: inaccuracy of the lab results, which originates from normal human errors or lack of experience of a person conducting diagnosis, and large processing times. Consequently, doctors and specialists are likely to provide improper prescriptions to patients. With the improvement of the computational power of computers, however, the whole diagnosis process can be automated. Several methods in literature have been proposed for this purpose. Most of these methods demand the availability of high-end microscopy imaging systems to generate reliable and accurate results. Such costly advanced devices may not be afforded by developing countries with sluggish economic growth. In this paper, therefore, we have developed a cost-effective framework which can address the mentioned challenge. Our approach introduces a Super Resolution (SR) model into the existing framework to enhance the resolution of the input images before letting them subjected to the subsequent detection stages. This provides a possibility for applying the low-end microscopy devices capable of capturing Low Resolution (LR) blood smear images for identifying the degree of Malaria in a patient. In the proposed framework, the SR component uses the nonlinear Charbonnier diffusion model in the regularization part because of its good regularity characteristics. Experimental results demonstrate strong correlation of our method and the manual one.
  • Thumbnail Image
    Item
    An Automatic and Cost-Effective Parasitemia Identification Framework for Low-End Microscopy Imaging Devices
    (IEEE, 2014) Maiseli, Baraka J.; Mei, Jiangyuan; Gao, Huijun; Yin, Shen
    In the detection of Malarial parasites from a patient, it is usually necessary to carefully examine the corresponding blood-slide smear and distinguish the infected and healthy Red Blood Cells (RBCs). If this process is done manually, as evidenced in common traditional approaches, the following challenges may be encountered: inaccuracy of the lab results, which originates from normal human errors or lack of experience of a person conducting diagnosis, and large processing times. Consequently, doctors and specialists are likely to provide improper prescriptions to patients. With the improvement of the computational power of computers, however, the whole diagnosis process can be automated. Several methods in literature have been proposed for this purpose. Most of these methods demand the availability of high-end microscopy imaging systems to generate reliable and accurate results. Such costly advanced devices may not be afforded by developing countries with sluggish economic growth. In this paper, therefore, we have developed a cost-effective framework which can address the mentioned challenge. Our approach introduces a Super Resolution (SR) model into the existing framework to enhance the resolution of the input images before letting them subjected to the subsequent detection stages. This provides a possibility for applying the low-end microscopy devices capable of capturing Low Resolution (LR) blood smear images for identifying the degree of Malaria in a patient. In the proposed framework, the SR component uses the nonlinear Charbonnier diffusion model in the regularization part because of its good regularity characteristics. Experimental results demonstrate strong correlation of our method and the manual one.
  • Thumbnail Image
    Item
    Edge Preservation Image Enlargement and Enhancement Method Based on the Adaptive Perona–Malik Non-Linear Diffusion Model
    (IEEE, 2014) Maiseli, Baraka J.; Elisha, Ogada Achieng; Mei, Jiangyuan; Gao, Huijun
    In this study, the authors have proposed a new super resolution (SR) model based on the Perona–Malik regularisation scheme. The new model integrates into its regularisation component an adaptive exponential term which automatically adjusts itself depending on the local image features. This lends more sensitivity and adaptability to the proposed model, thereby making the reconstruction process much less punishing against semantically important features. Therefore, regularisation is stronger in homogeneous regions, and weaker in the neighbourhood of boundaries. The proposed method has a promising capability of supressing noise more effectively, while preserving important image features. The approach used differs significantly from the available methods, especially in the manner in which adaptability has been deployed. Noting that SR methods are less sensitive to the local image topography, a factor that causes the super-resolved images to be visually poor, the new method sensitively probes the local features of the image, and determines the necessary level of reconstruction and regularisation. Additionally, the formulation robustly introduces a backward diffusion, a phenomenon proved from literature to have a tendency of sharpening edges. The authors have included empirical reconstruction results to demonstrate that their model produces better images in comparison with other classical methods.
  • Thumbnail Image
    Item
    Lp-TV model for structure extraction with end-to-end contour learning
    (IEEE, 2017) Song, Chunwei; Maiseli, Baraka J.; Zuo, Wangmeng; Gao, Huijun
    Structure extraction is important for human perception. However, for various textured images, computers can hardly achieve this goal. Despite a plethora of studies to address the challenge, results from most previous methods contain unwanted artifacts and over-smoothed structures. Therefore, to address the weaknesses, we have proposed a variational model with end-to-end contour learning capability. Our formulation dwells in two observations: likelihood for representation of residual textures may be well abstracted using super Gaussian distribution, and edge metrics with semantic meaning may benefit structure preservation. The augmented Lagrangian method is adopted for optimal computation. Compared with classical approaches, our method offers a higher performance in structure extraction, including situations where the images have significant nonuniformity of the scale features.
  • Thumbnail Image
    Item
    A multi-frame super-resolution method based on the variable-exponent nonlinear diffusion regularizer
    (EURASIP Journal on Image and Video Processing, 2015-07-28) Maiseli, Baraka J.; Elisha, Ogada Achieng; Gao, Huijun
    In this work, the authors have proposed a multi-frame super-resolution method that is based on the diffusion-driven regularization functional. The new regularizer contains a variable exponent that adaptively regulates its diffusion mechanism depending upon the local image features. In smooth regions, the method favors linear isotropic diffusion, which removes noise more effectively and avoids unwanted artifacts (blocking and staircasing). Near edges and contours, diffusion adaptively and significantly diminishes, and since noise is hardly visible in these regions, an image becomes sharper and resolute—with noise being largely reduced in flat regions. Empirical results from both simulated and real experiments demonstrate that our method outperforms some of the state-of-the-art classical methods based on the total variation framework.
  • Thumbnail Image
    Item
    A multi-frame super-resolution method based on the variable-exponent nonlinear diffusion regularizer
    (2015-07) Maiseli, Baraka; Ogada, Elisha A.; Gao, Huijun
    In this work, the authors have proposed a multi-frame super-resolution method that is based on the diffusion-driven regularization functional. The new regularizer contains a variable exponent that adaptively regulates its diffusion mechanism depending upon the local image features. In smooth regions, the method favors linear isotropic diffusion, which removes noise more effectively and avoids unwanted artifacts (blocking and staircasing). Near edges and contours, diffusion adaptively and significantly diminishes, and since noise is hardly visible in these regions, an image becomes sharper and resolute—with noise being largely reduced in flat regions. Empirical results from both simulated and real experiments demonstrate that our method outperforms some of the state-of-the-art classical methods based on the total variation framework
  • Thumbnail Image
    Item
    A Noise-Suppressing and Edge-Preserving Multiframe Super-Resolution Image Reconstruction Method
    (Elsevier, 2015) Maiseli, Baraka J.; Ally, Nassor; Gao, Huijun
    Super-resolution technology is an approach that helps to restore high quality images and videos from degraded ones. The method stems from an ill-posed minimization problem, which is usually solved using the L2 norm and some regularization techniques. Most of the classical regularizing functionals are based on the Total Variation and the Perona–Malik frameworks, which suffer from undesirable artifacts (blocking and staircasing). To address these problems, we have developed a super-resolution framework that integrates an adaptive diffusion-based regularizer. The model is feature-dependent: linear isotropic in flat regions, a condition that regularizes an image uniformly and removes noise more effectively; and nonlinear anisotropic near boundaries, which helps to preserve important image features, such as edges and contours. Additionally, the new regularizing kernel incorporates a shape-defining parameter that can be automatically updated to ensure convexity and stability of the corresponding energy functional. Comparisons with other methods show that our method is superior and, more importantly, can achieve higher reconstruction factors.
  • Thumbnail Image
    Item
    A noise-suppressing and edge-preserving multiframe super-resolution image reconstruction method
    (Signal Processing: Image Communication, 2015-03-12) Maiseli, Baraka J.; Ally, Nassor; Gao, Huijun
    Super-resolution technology is an approach that helps to restore high quality images and videos from degraded ones. The method stems from an ill-posed minimization problem, which is usually solved using the L2 norm and some regularization techniques. Most of the classical regularizing functionals are based on the Total Variation and the Perona–Malik frameworks, which suffer from undesirable artifacts (blocking and staircasing). To address these problems, we have developed a super-resolution framework that integrates an adaptive diffusion-based regularizer. The model is feature-dependent: linear isotropic in flat regions, a condition that regularizes an image uniformly and removes noise more effectively; and nonlinear anisotropic near boundaries, which helps to preserve important image features, such as edges and contours. Additionally, the new regularizing kernel incorporates a shape-defining parameter that can be automatically updated to ensure convexity and stability of the corresponding energy functional. Comparisons with other methods show that our method is superior and, more importantly, can achieve higher reconstruction factors.
  • Thumbnail Image
    Item
    Recent developments and trends in point set registration methods
    (Journal of Visual Communication and Image Representation, 2017) Maiseli, Baraka J.; Gu, Yanfeng; Gao, Huijun
    Point set registration (PSR) is the process of computing a spatial transformation that optimally aligns pairs of point sets. The method helps to amalgamate multiple datasets into a common coordinate system. Because of their immense practical applications, several studies have attempted to address challenges inherent in the PSR problem. However, limited works exist to discuss recent developments, failures, and trends of the PSR methods. To date, a classical work of Tam et al., published in 2013, can be regarded as a comprehensive review paper for registration methods. Nevertheless, this work has inadequately revealed a range of possible knowledge gaps of the previous studies. Additionally, since the publication year of their work, more superior and state-of-the-art methods have been proposed. The present study surveys PSR approaches until 2017, and our primary focus is to expose central ideas and limitations of the methods to facilitate experts and practitioners advance the field.
  • Thumbnail Image
    Item
    Robust cost function for optimizing chamfer masks
    (The Visual Computer, 2018) Maiseli, Baraka J.; Bai, LiFei; Yang, Xianqiang; Gu, Yanfeng; Gao, Huijun
    Chamfering, a mask-driven technique, refers to a process of propagating local distances over an image to estimate a reference metric. Performance of the technique depends on the design of chamfer masks using cost functions. To date, most scholars have been using a mean absolute error and a mean squared error to formulate optimization problems for estimating weights in the chamfer masks. However, studies have shown that these optimization functions endure some potential weaknesses, including biasedness and sensitivity to outliers. Motivated by the weaknesses, the present work proposes an alternative difference function, RLog, that is unbiased, symmetrical, and robust. RLog takes the absolute logarithm of the relative accuracy of the estimated distance to compute optimal chamfer weights. Also, we have proposed an algorithm to map entries of the designed real-valued chamfer masks into integers. Analytical and experimental results demonstrate that chamfering based on our weights generate polygons and distance maps with lower errors. Methods and results of our work may be useful in robotics to address the matching problem.
  • Thumbnail Image
    Item
    Robust edge detector based on anisotropic diffusion-driven process
    (Information Processing Letters, 2016-05-01) Maiseli, Baraka J.; Gao, Huijun
    Edge detection involves a process to discriminate, highlight, and extract useful image features (edges and contours). In many situations, we prefer an edge detector that distinguishes these features more accurately, and which comfortably deals with a variety of data. Our observations, however, discovered that most edge-defining functionals underperform and generate false edges under poor imaging conditions. Therefore, the current research proposes a robust diffusion-driven edge detector for seriously degraded images. The method is iterative, and suppresses noise while simultaneously marking real edges and deemphasizing false edges. The anisotropic nature of the new functional helps to remove noise and to preserve semantic structures. Even more importantly, the functional exhibits a forward–backward behavior that may sharpen and strengthen edges. Comparisons with some other classical approaches demonstrate superiority of the proposed approach.
  • Thumbnail Image
    Item
    Robust edge detector based on anisotropic diffusion-driven process
    (2015-12) Maiseli, Baraka; Gao, Huijun
    Edge detection involves a process to discriminate, highlight, and extract useful image features (edges and contours). In many situations, we prefer an edge detector that distinguishes these features more accurately, and which comfortably deals with a variety of data. Our observations, however, discovered that most edge-defining functionals underperform and generate false edges under poor imaging conditions. Therefore, the current research proposes a robust diffusion-driven edge detector for seriously degraded images. The method is iterative, and suppresses noise while simultaneously marking real edges and deemphasizing false edges. The anisotropic nature of the new functional helps to remove noise and to preserve semantic structures. Even more importantly, the functional exhibits a forward–backward behavior that may sharpen and strengthen edges. Comparisons with some other classical approaches demonstrate superiority of the proposed approach.
  • Thumbnail Image
    Item
    A robust super-resolution method with improved high-frequency components estimation and aliasing correction capabilities
    (Journal of the Franklin Institute, 2014-01) Maiseli, Baraka J.; Wu, Chuan; Mei, Jiangyuan; Liu, Qiang; Gao, Huijun
    In this paper, we have proposed a robust super-resolution high-frequency component estimation (RS-HFCE) method, which can efficiently estimate lost high-frequency components and correct aliasing effects of low-frequency components of an image. The fundamental principle of operation of the proposed method is based on the idea that, when a baseband band-limited image signal of known bandwidth in a high-resolution lattice is iteratively low-pass filtered in the frequency domain, the unknown values in the lattice can be interpolated, thus correcting the aliasing for the low-frequency components. If this process is done along with adjusting the amplitudes of the known pixel values, some high-frequency components of an image are automatically extrapolated. In order to provide simultaneous edge preservation and noise removal capabilities of the super-resolved images, an improved version of an adaptive Perona–Malik (PM) model is incorporated into the process. One of the characteristics of the proposed method is its high level of tolerance capabilities to reconstruction errors and noise caused by an increase in the reconstruction scaling factors. High quality images of higher resolution are still appreciably reconstructed when greater magnification factors are used. From a couple of experiments on real images, and using both subjective and objective image quality assessment measures, it is demonstrated that the proposed method outperforms most of other classical methods.