Information Of Input Image Research Articles

Forming Optimal Projection Images from Intra-Retinal Layers Using Curvelet-Based Image Fusion Method.

Background:Image fusion is the process of combining the information of several input images into one image. Projection images obtained from three-dimensional (3D) optical coherence tomography (OCT) can show inlier retinal pathology and abnormalities that are not visible in conventional fundus images. In recent years, the projection image is often made by an average on all retina that causes to lose many intraretinal details.Methods:In this study, we focus on the formation of optimum projection images from retinal layers using Curvelet-based image fusion. The latter consists of three main steps. In the earlier studies, macular spectral 3D data using diffusion map-based OCT were segmented into 12 different boundaries identifying 11 retinal layers in three dimensions. In the second step, projection images are attained using conducting some statistical methods on the space between each pair of boundaries. In the next step, retinal layers are merged using Curvelet transform to make the final projection images.Results:These images contain integrated retinal depth information as well as an ideal opportunity to better extract retinal features such as vessels and the macula region. Finally, qualitative and quantitative evaluations show the superiority of this method to the average-based and wavelet-based fusion methods. Overall, our method obtains the best results for image fusion in all terms such as entropy (6.7744) and AG (9.5491).Conclusion:Creating an image with more and detailed information made by the Curvelet-based image fusion has significantly higher contrast. There are also many thin veins in Curvelet-based fused image, which are absent in average-based and wavelet-based fused images.

Multiscale Feature Extraction Based on Convolutional Sparse Decomposition for Hyperspectral Image Classification

Due to the different spatial properties presented by various ground objects in hyperspectral image (HSI), multiscale-based feature extraction approaches have been developed for HSI classification in recent years. However, the spatial features of different scales are usually acquired at the cost of obscuring the structural information of input image, which severely limits the effectiveness of multiscale strategy. In this article, a convolutional sparse decomposition (CSD) model is introduced to characterize the significant spatial structures of hyperspectral data while removing the irrelevant noise and local textures at the specific scale. Based on the CSD model, a multiscale spectral-spatial feature extraction framework is generated, which consists of the following steps. First, the spectral dimensionality of the original HSI is reduced through a segmented averaging approach. Second, spatial features at different scales are separated from the dimension-reduced data by solving the CSD model with different regularization parameters. Finally, principal component analysis is performed and the obtained multiscale spectral-spatial features are stacked together for classification. Experiments conducted on three widely used hyperspectral datasets demonstrate that the proposed method is robust in capturing effective features of ground objects at different scales and leads to better classification results than several state-of-the-art methods.

Multiscale Convolutional Fusion Network for Non-Lambertian Photometric Stereo

One of the key issues in photometric stereo is the extension of its application in real world objects which shows non-Lambertian reflectance. This letter proposes a multi-scale weighted convolutional fusion network with deep learning architecture to realize high-precision perception of non-Lambertian surfaces under arbitrary illumination conditions. A multi-scale convolutional fusion module is designed to strengthen the photometric physics and the utilization of the neighborhood features at the same time so as to overcome shadows and distinguish multiple materials. In order to further deal with the problem of arbitrary illumination conditions, a multi-resolution polar coordinate division method is proposed to integrate the input image information and fully utilizes the multi-scale convolution. Both syntheses and real-world experiments verifies the performance of the proposed method in recovery accuracy and computational efficiency.

Super resolution land cover mapping of hyperspectral images using the deep image prior-based approach

ABSTRACTDue to the instantaneous field-of-view (IFOV) of the sensor and diversity of land cover types, some pixels, usually named mixed pixels, contain more than one land cover type. Soft classification can predict the portion of each land cover type in mixed pixels in the absence of spatial distribution. The spatial distribution information in mixed pixels can be solved by super resolution mapping (SRM). Typically, SRM involves two steps: soft class value estimation, which is similar to the image super resolution of image restoration, and land cover allocation. A new SRM approach utilizes a deep image prior (DIP) strategy combined with a super resolution convolutional neural network (SRCNN) to estimate fine resolution fraction images for each land cover type; then, a simple and efficient classifier is used to allocate subpixel land cover types under the constraint of the generated fine fraction images. The proposed approach can use prior information of input images to update network parameters and no longer require training data. Experiments on three different cases demonstrate that the subpixel classification accuracy of the proposed DIP-based SRM approach is significantly better than the three conventional SRM approaches and a transfer learning-based neural network SRM approach. In addition, the DIP-SRM approach performs very robustly about small-area objects within multiple land cover types and significantly reduces soft classification uncertainty. The results of this paper provide an extension for utilizing SRCNN to address SRM issues in hyperspectral images.

Dilated Skip Convolution for Facial Landmark Detection.

Facial landmark detection has gained enormous interest for face-related applications due to its success in facial analysis tasks such as facial recognition, cartoon generation, face tracking and facial expression analysis. Many studies have been proposed and implemented to deal with the challenging problems of localizing facial landmarks from given images, including large appearance variations and partial occlusion. Studies have differed in the way they use the facial appearances and shape information of input images. In our work, we consider facial information within both global and local contexts. We aim to obtain local pixel-level accuracy for local-context information in the first stage and integrate this with knowledge of spatial relationships between each key point in a whole image for global-context information in the second stage. Thus, the pipeline of our architecture consists of two main components: (1) a deep network for local-context subnet that generates detection heatmaps via fully convolutional DenseNets with additional kernel convolution filters and (2) a dilated skip convolution subnet—a combination of dilated convolutions and skip-connections networks—that are in charge of robustly refining the local appearance heatmaps. Through this proposed architecture, we demonstrate that our approach achieves state-of-the-art performance on challenging datasets—including LFPW, HELEN, 300W and AFLW2000-3D—by leveraging fully convolutional DenseNets, skip-connections and dilated convolution architecture without further post-processing.

Image based Road Surface Classification Method using CNN

Road detection and road surface classification in autonomous driving are the most basic and important issues. In this paper, we propose a data augmentation method for road surface classification using image information. We design an optimal network that can classify the type of road surface from the input image information and propose a data increase technique that can efficiently judge by using limited data to improve learning performance. To verify the proposed methods, many running images were used on the Internet. Experimental vehicle was developed and applied to verify the developed networks and it shows that they operate accurately in real time.

Image Fusion Using Adjustable Non-subsampled Shearlet Transform

Multiscale transform (MST) and sparse representation (SR)-based image fusion techniques are widely used in many applications. However, the existing fusion methods are inefficient to perfectly seize significant details and texture information of input images because of some limitations of MST-and SR-based techniques. Moreover, it is quite important to measure singularities and structural information of the fused image to objectively measure the performance of different fusion techniques. Therefore, a metric which can efficiently measure singularity and structural information of the fused image would be quite useful for comparison of different fusion techniques. To address these issues, a modified Meyer window-based adjustable nonsubsampled shearlet transform (ANSST) is proposed for decomposition of preregistered input images into low-and high-frequency coefficients. The low-frequency bands are fused by convolutional SR modeling. The high-frequency bands are fused by our proposed information entropy, standard deviation, and range descriptor, which considers entropy, standard deviation, and range-filtering features. Moreover, a metric $Q_{SS}$ is formulated to measure the singularities and structural information of the fused image. The $Q_{SS}$ metric is developed by using the concept of the proposed native division and native difference filtering. To show the applicability of the proposed image framework, the proposed ANSST is applied for image denoising. Experimental results verify the effectiveness of the proposed fusion method and the metric $Q_{SS}$ .

Image denoising via deep residual convolutional neural networks

Recently, convolutional neural network (CNN)-based methods have achieved impressive performance on image denoising. Notably, CNN with deeper and thinner structures is more flexible to extract the image details. However, direct stacking some existing networks is difficult to achieve satisfactory denoising performance. In this paper, we propose a novel deep residual convolutional neural network (DRCNN) for image denoising. The main structure of DRCNN is the residual block that consists of two convolutional layers, and there are skip connections between these two convolutional layers without the batch normalization operation. The skip connection not only directly transfers the input image information to the hidden layer but also reduces the path length of gradient transfer, making the gradient transfer in a short path and alleviating the vanishing-gradient problem. DRCNN is compared with several state-of-the-art algorithms, and the experimental results demonstrated its denoising effectiveness.

Image Captioning with Memorized Knowledge

Image captioning, which aims to automatically generate text description of given images, has received much attention from researchers. Most existing approaches adopt a recurrent neural network (RNN) as a decoder to generate captions conditioned on the input image information. However, traditional RNNs deal with the sequence in a recurrent way, squeezing the information of all previous words into hidden cells and updating the context information by fusing the hidden states with the current word information. This may miss the rich knowledge too far in the past. In this paper, we propose a memory-enhanced captioning model for image captioning. We firstly introduce an external memory to store the past knowledge, i.e., all the information of generated words. When predicting the next word, the decoder can retrieve knowledge information about the past by means of a selective reading mechanism. Furthermore, to better explore the knowledge stored in the memory, we introduce several variants that consider different types of past knowledge. To verify the effectiveness of the proposed model, we conduct extensive experiments and comparisons on the well-known image captioning dataset MS COCO. Compared with the state-of-the-art captioning models, the proposed memory-enhanced captioning model shows a significant improvement in terms of the performance (improving 3.5% in terms of CIDEr). The proposed memory-enhanced captioning model, as demonstrated in the experiments, is more effective and superior to the state-of-the-art methods.

Review of State-of-Art Image Restoration Techniques

In recent years, image processing is an essential technique for most of the applications such as business, government sectors, medial industry, cinema field, and etc. In image processing, most of the subdomains are there which helps to do their respective process. Among that, Image Restoration (IR) is one of the important domain which helps to retrieve the original image information from the noisy image. Normally, input image is affected by noise during the transmission. So, noise need to be removed to get the original input image information. In this paper, different kind of IR techniques are explained which is implemented in Field Programmable Gate Array (FPGA). In VLSI domain, FPGA is the important process which is used to identify the hardware utilization of the entire algorithm. The process and different methods of IR is studied in this paper.

Global and Local Oriented Gabor Texture Histogram for Person Re-identification

ABSTRACT: Automated person re-identification is a key process in global distributed camera systems. This paper proposes a new feature, the Global and Local-Oriented Gabor Texture Histogram (GLOGTH), for person re-identification. GLOGTH is a combination of the local texture and global structure information of a given input image. This feature aims at representing the human appearance traits with low-dimensional feature extraction. The proposed feature extracts the texture information of input images based on the orientation of the weighted gradient from the global representation. In GLOGTH, the principal orientation is determined by the gradient of the pixels. Based on the principal orientation, the Gabor feature is extracted and imbues GLOGTH with the strong ability to express edge information, apart from making it robust to lighting variances. The experimental results acquired from the databases demonstrate that the proposed GLOGTH framework is capable of achieving notable improvements, in many cases reaching higher classification accuracy than traditional frameworks.

Recent Medical Image Fusion Techniques: A Review

Image fusion technique has been widely in use in many applications such as computer vision, surveillance, robotics, medical imaging, remote sensing etc. Medical image fusion plays a vital role to deal with medical issues reflected through images of human body, organs and cells. It is necessary to preserve the fused image in order to keep the relevant information of input images. It is also necessary to keep the fusion process away from introducing any artifacts or inconsistencies. More over image fusion techniques must provide a good quality and must enhance the applications of recent data. This paper elucidates the survey of recent literatures, which dealt medical images like magnetic resonance imaging (MRI), Computed Tomography (CT), Single Photon Emission Computed Tomography (SPECT) and Positron Emission Tomography (PET) of brain. This paper reviews some of the recent medical image fusion algorithm such as Non Sampled Contourlet transform, Directive Contrast, Daubechies complex wavelet transform, Shift-invariant Shearlet transform etc. The merits and demerits of the various image fusion algorithm which developed by the authors is described in the paper. The survey emerges out the optimum fusion algorithm.

Anfis to Detect Brain Tumor Using MRI

Processing of Magnetic Resonance Imaging(MRI) is one of the widely known best techniques to diagnose brain tumor since it gives better results than ultrasound or X-Ray images. The main objective is to diagnose the presence and extraction of brain tumor using MRI images. Image preprocessing includes contrast stretching, noise filtering and Adaptive Histogram Equalization(AHE). AHE gives a graphical representation of digital image without enhancing above the desired level. The next stage involves transferring the redundant information in input image to reduced set of features is called feature selection and is done by color, shape or texture of an image. Image is segmented using incorporation of Artificial Neural Networks(ANN) and Fuzzy logic called Adaptive Neuro-Fuzzy Inference System(ANFIS) wherein we get the desired output to differentiate tumor affected and normal image with its severity level. Since we deal with uncertainty much more, fuzzy logic serves as a vibrant tool in representing human knowledge as IF-THEN rules. MATLAB has been implemented in detection and extraction of tumor at an early stage.

Optimum Wavelet-Based Homomorphic Medical Image Fusion Using Hybrid Genetic–Grey Wolf Optimization Algorithm

Medical image fusion techniques have been widely used in various clinical applications. Generalized homomorphic filters have Fourier domain features of input image. In multimodal medical image fusion discrete wavelet transform-based techniques provides more features and is performed over Fourier spectrum. In this paper, we proposed a homomorphic wavelet fusion which is called optimum homomorphic wavelet fusion (OHWF) using hybrid genetic–grey wolf optimization (HG-GWO) Algorithm. In OHWF, which consists of logarithmic and wavelet domain information of input images. The wavelet-based homomorphic fusion consists of multilevel decomposition features of input image. In our proposal, the approximation coefficients of modality1 (anatomical structure) and optimum scaled detailed coefficients of modality2 are given to adder1. In adder 2, the optimum scaled detailed coefficients of modality 1 and approximation coefficients of modality 2 are added together. The resultants of adder 1 and adder 2 are fused together using pixel based averaging rule. First, the proposed fusion approach is validated for MR-SPECT, MR-PET, MR-CT, and MR T1-T2 image fusion using various fusion evaluation indexes. Later, the conventional grey wolf optimization is modified with genetic operator. Experimental results show that the proposed approach outperforms state-of-the-art fusion algorithms in terms of both structural and the functional information in the fused image.

Visibility distance estimation in foggy situations and single image dehazing based on transmission computation model

The existing visibility distance estimation algorithms in foggy situations use the region growing method to extract the vertical position of inflection point of image intensity changing. These algorithms have lower inflection point location accuracy for an image with a non-homogeneous road surface. To deal with these problems, this study presents a novel visibility distance measuring technique under foggy weather conditions. This method combines two major models: inflection point estimation (IPE) model and transmission refining (TR) model. The proposed IPE model based on transmission computation model derives a very useful relation between the transmission value of inflection points and the constant ${\rm e}^{ - 2}$e -2 . In order to acquire the more accurate transmission map and vertical position of each inflection point, this study establishes an effective TR model. This model exploits the edge information of input images, in order to significantly reduce the effects of artefact. The proposed algorithm provides more accurate visibility distance estimation of an image with a non-homogeneous road surface than the well-known algorithm through qualitative evaluations in experiments. The experimental results also show that the TR model has better outcomes than the guided filter approach through qualitative and quantitative evaluations.

A self-adaptive multi-objective harmony search based fuzzy clustering technique for image segmentation

Image segmentation can be considered as a problem of clustering since the pixels in the digital image are clustered in term of some evaluation criteria. Generally, clustering technique in image segmentation employs a single objective which can not reach ideal result for various kinds of images. Moreover, fuzzy c-means (FCM) algorithms which determine the fuzzy partition matrix of the data set by solving the clustering problem with conditional constraints and obtain the clustering output, have been verified effective and efficient for image segmentation. In fact, these FCM algorithms still have some shortcomings including: being sensitive to outliers and noise, key parameters need to be adjusted with experience. In view of this, a self-adaptive multi-objective harmony search based fuzzy clustering (SAMOHSFC) technique for image segmentation is proposed in this paper. SAMOHSFC technique encodes several cluster centers in one harmony vector and optimizes multiple objectives. In addition, we consider the spatial information of the image as an attribute of the input data set besides the attribute of gray information of input image in the SAMOHSFC. Superiority of the proposed algorithm over three classic segmentation algorithms has been verified for a synthetic and two real images from quantitative and visual aspect. In the experiment, the effect of different kinds of spatial information on the segmentation performance of the SAMOHSFC is analyzed.

Multi-modality image fusion based on enhanced fuzzy radial basis function neural networks

In clinical applications, single modality images do not provide sufficient diagnostic information. Therefore, it is necessary to combine the advantages or complementarities of different modalities of images. Recently, neural network technique was applied to medical image fusion by many researchers, but there are still many deficiencies. In this study, we propose a novel fusion method to combine multi-modality medical images based on the enhanced fuzzy radial basis function neural network (Fuzzy-RBFNN), which includes five layers: input, fuzzy partition, front combination, inference, and output. Moreover, we propose a hybrid of the gravitational search algorithm (GSA) and error back propagation algorithm (EBPA) to train the network to update the parameters of the network. Two different patterns of images are used as inputs of the neural network, and the output is the fused image. A comparison with the conventional fusion methods and another neural network method through subjective observation and objective evaluation indexes reveals that the proposed method effectively synthesized the information of input images and achieved better results. Meanwhile, we also trained the network by using the EBPA and GSA, individually. The results reveal that the EBPGSA not only outperformed both EBPA and GSA, but also trained the neural network more accurately by analyzing the same evaluation indexes.

Fusion of Color Images Based on Fuzzy Transform and Spatial Frequency

Multiple images of a scene, captured using different imaging devices and containing complementary information, are required to be combined into a single fused image. The fused image, so obtained, should ideally contain all important features of individual input images. Further, color image processing is also gaining importance in most of the real-life applications. Color images provide better visual information and are highly suitable for human visual perception. Motivated by the gaining importance of color image processing, fusion of color images is proposed. The fusion algorithm is proposed in HSV color space where the luminance components of multiple input images, based on fuzzy transform and spatial frequency, are fused. The chrominance information of input images provides color to the fused image. The fused image contains all important information and possesses natural color appearance. Experiments performed on various pairs of visible–infrared and multifocus images demonstrate the superiority of the proposed algorithm over recent state-of-the-art image fusion algorithms.

Research on Convolutional Neural Network Model for Sonar IMAGE Segmentation

The speckle noise of sonar images affects the human interpretation and automatic recognition of images seriously. It is important and difficult to realize the precision segmentation of sonar image with speckle noise in the field of image processing. Full convolution neural network (FCN) has the advantage of accepting arbitrary size image and preserving spatial information of original input image. In this paper, the image features are obtained by autonomic learning of convolutional neural network, the original learning rules based on the mean square error loss function is improved. Taking the pixel as the processing unit, the segmentation method based on FCN model with relative loss function(FCN-RLF) for small submarine sonar image is proposed, sonar image pixel-level segmentation is achievied. Experimental results show that the improved algorithm can improve the segmentation accuracy and keep the edge and detail of sonar image better. The proposed model has better ability to reject sonar image speckle noise.

An Integrated Dictionary-Learning Entropy-Based Medical Image Fusion Framework

Image fusion is widely used in different areas and can integrate complementary and relevant information of source images captured by multiple sensors into a unitary synthetic image. Medical image fusion, as an important image fusion application, can extract the details of multiple images from different imaging modalities and combine them into an image that contains complete and non-redundant information for increasing the accuracy of medical diagnosis and assessment. The quality of the fused image directly affects medical diagnosis and assessment. However, existing solutions have some drawbacks in contrast, sharpness, brightness, blur and details. This paper proposes an integrated dictionary-learning and entropy-based medical image-fusion framework that consists of three steps. First, the input image information is decomposed into low-frequency and high-frequency components by using a Gaussian filter. Second, low-frequency components are fused by weighted average algorithm and high-frequency components are fused by the dictionary-learning based algorithm. In the dictionary-learning process of high-frequency components, an entropy-based algorithm is used for informative blocks selection. Third, the fused low-frequency and high-frequency components are combined to obtain the final fusion results. The results and analyses of comparative experiments demonstrate that the proposed medical image fusion framework has better performance than existing solutions.