检索结果-内蒙古大学图书馆

Brain tumour detection in magnetic resonance imaging using Levenberg-Marquardt backpropagation neural network

IET image processing 2023年第1期17卷 88-103页

作者： Ghahramani, Marzieh Shiri, Nabiollah Islamic Azad Univ Shiraz Branch Dept Elect Engn Shiraz *** Iran

Magnetic resonance imaging (MRI) is a high-quality medical image that is used to detect brain tumours in a complex and time-consuming manner. In this study, a back propagation neural network (BPNN) along with the Levenberg-Marquardt algorithm (LMA) is proposed to classify MRIs and diagnose brain tumours in a simple and fast process. The BPNN has 10 neurons in the hidden layer, and the default function of the feedforward feeds is mean squared error (MSE). The LMA is optimized as a multivariable adaptive approach and considerably decreases the MSE of the BPNN, so the errors of the tumour classification are diminished. The proposed method follows four steps including preprocessing, skull removal, feature extraction, and classification. The input MRIs are converted to greyscale, resized, and thresholding is performed in the preprocessing step and followed by skull removal. Morphological operations of closing, opening, and dilation are used to segment abnormal areas in the MRIs, and the opening operator recognizes the tumour more accurately. Using statistical analysis and a grey-level co-occurrence matrix (GLCM) 12 features are extracted from the MRIs and used as the inputs of the BPNN. To evaluate the proposed method, 670 normal and 670 abnormal brain MRIs are used as input data, and the classification is performed in 0.494 s. The accuracy, sensitivity, specificity, precision, dice, recall, and MSE are 98.7%, 97.61%, 99.7%, 97.61%, 98.6%, 97.61%, and 0.005, respectively. The approach is accurate and fast for medical images classification.

关键词： medical image classification image segmentation feature extraction tumour classification default function Biology and medical computing input MRIs high-quality medical image image classification biomedical MRI Medical magnetic resonance imaging and spectroscopy multivariable adaptive approach brain tumour detection Biophysics of neurophysiological processes brain tumour diagnosis neural nets backpropagation mean squared error opening operator greyscale Computer vision and image processing techniques Biomedical magnetic resonance imaging and spectroscopy statistical analysis grey-level cooccurrence matrix neural nets skull removal hidden layer preprocessing step image recognition MSE preprocessing removal Levenberg-Marquardt algorithm magnetic resonance imaging image thresholding abnormal brain MRIs BPNN morphological operations brain LMA Other topics in statistics Patient diagnostic methods and instrumentation tumours Levenberg-Marquardt backpropagation neural network medical image processing Probability theory, stochastic processes, and statistics

来源：评论

学校读者我要写书评

暂无评论

Improve the efficiency of handcrafted features in image retrieval by adding selected feature generating layers of deep convolutional neural networks

引用

signal image AND VIDEO processing 2024年第3期18卷 2607-2620页

作者： Shamsipour, Ghazal Fekri-Ershad, Shervan Sharifi, Mahdi Alaei, Alireza Islamic Azad Univ Fac Comp Engn Najafabad Branch Najafabad Iran Islamic Azad Univ Big Data Res Ctr Najafabad Branch Najafabad Iran Southern Cross Univ Fac Sci & Engn East Lismore NSW Australia

Today, with the rapid growth of communication technology and the development of social networks and smartphones, the amount of data stored by users in the form of images has significantly increased. Therefore, several researchers have focused on the field of image retrieval in the past decade. image retrieval means retrieving the most similar images to the query sample from a vast image database in terms of content and semantic. So far, various image retrieval methods based on feature engineering or deep features have been proposed. Most methods that use handcrafted features, despite their lower runtime, usually provide lower performance than deep learning-based approaches. In this paper, to increase the efficiency of handcrafted features in image retrieval, their combination with the output of residual blocks in deep neural networks is used in the feature fusion phase. In this paper, the efficiency of feature generation layers in common deep networks such as residual, classical sequential convolution and bottleneck has been analyzed in order to increase the efficiency of image retrieval systems based on handcrafted features. For this purpose, in most popular deep networks, the layers that perform the classification task are removed, and the output of feature generation layers at different depths, with the help of a flattened layer, is converted to feature vectors that can be connected to handcrafted made numerical features in retrieval systems. A variety of popular handcrafted features in the spatial domain, such as color and texture and wavelet in frequency, are employed to generate handcrafted features. This paper presents a novel hybrid feature set based on a combination of classical feature engineering techniques and deep convolutional neural networks suitable for image retrieval. The efficiency of the proposed method is evaluated in terms of precision and recall metrics on the benchmark datasets, such as Corel-1 and -5k. The proposed method provided an accurac

关键词： image retrieval Feature extraction Deep learning Texture analysis Residual neural network

来源：评论

学校读者我要写书评

暂无评论

Real-Time Attentive Dilated U-Net for Extremely Dark image Enhancement

引用

ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS 2024年第8期20卷 1-19页

作者： Huang, Junjian Ren, Hao Liu, Shulin Liu, Yong Lv, Chuanlu Lu, Jiawen Xie, Changyong Lu, Hong Fudan Univ Sch Comp Sci Shanghai Key Lab Intelligent Informat Proc Shanghai 200438 Peoples R China Naval Med Univ Naval Med Ctr Dept Aviat Med Shanghai 200433 Peoples R China Naval Med Univ Naval Med Ctr Dept Med Equipment Shanghai 200433 Peoples R China Naval Med Univ Naval Med Ctr Dept Med Res Shanghai 200433 Peoples R China Naval Med Univ Sch Basic Med Sci Shanghai 200433 Peoples R China

images taken under low-light conditions suffer from poor visibility, color distortion, and graininess, all of which degrade the image quality and hamper the performance of downstream vision tasks, such as object detection and instance segmentation in the field of autonomous driving, making low-light enhancement an indispensable basic component of high-level visual tasks. Low-light enhancement aims to mitigate these issues, and has garnered extensive attention and research over several decades. The primary challenge in low-light image enhancement arises from the low signal-to-noise ratio caused by insufficient lighting. This challenge becomes even more pronounced in near-zero lux conditions, where noise overwhelms the available image information. Both traditional image signal processing pipeline and conventional low-light image enhancement methods struggle in such scenarios. Recently, deep neural networks have been used to address this challenge. These networks take unmodified RAW images as input and produce the enhanced sRGB images, forming a deep learning based image signal processing pipeline. However, most of these networks are computationally expensive and thus far from practical use. In this article, we propose a lightweight model called attentive dilated U-Net (ADU-Net) to tackle this issue. Our model incorporates several innovative designs, including an asymmetric U-shape architecture, dilated residual modules for feature extraction, and attentive fusion modules for feature fusion. The dilated residual modules provide strong representative capability, whereas the attentive fusion modules effectively leverage low-level texture information and high-level semantic information within the network. Both modules employ a lightweight design but offer significant performance gains. Extensive experiments demonstrate that our method is highly effective, achieving an excellent balance between image quality and computational complexity-that is, taking less than 4ms for a

关键词： Low-light image enhancement image signal processing convolutional neural networks attention mechanism

来源：评论

学校读者我要写书评

暂无评论

IoT-based heart disease risk prediction and categorization for healthcare using an optimized progressive residual attention-based holographic convolutional neural network

引用

signal image AND VIDEO processing 2025年第3期19卷 1-17页

作者： Sreenivasulu, G. Anithaashri, T. P. Saveetha Inst Med & Tech Sci Saveetha Sch Engn Dept Comp Sci & Engn Chennai India

The rapid growth of smart healthcare monitoring equipment can be attributed to the emergence of portable medical gadgets that possess Internet of Things (IoT) capabilities. Disease prevention is aided by the shift in emphasis from traditional in-person consultations to telemedicine brought about by the incorporation of IoT and deep learning (DL) in healthcare. Nevertheless, there are several limitations with the current methods when it comes to informing patients of their health state based on ongoing changes in their health for prompt healing. This study presents a brand-new smart healthcare monitoring system called Progressive Residual Attention Holographic Convolutional neural Network (PRAHCNN) based on Holographic Convolutional neural Network (HCNN) integrated with a Progressive Residual Attention Network (PRAN). By detecting the risk factors, the proposed study aims to give an accurate and timely diagnosis of cardiac illnesses. To assess the physiological data in this case, the suggested model continually gathers input data from wearable sensors and uses a neural network to process it. The risk status for cardiac disease is then divided into four categories by the system: low, medium, high, and extreme. Here, an automated telephone call and/or SMS notification, including the patient's location, is delivered to the patient's relative if the patient's condition is deemed to be at high or extreme risk. According to the experimental results, the suggested method uses the Cleveland database to obtain a better accuracy of 99.8% with a lower error rate of 2%, and the Hungarian dataset to reach a higher accuracy of 99.6% and a lower error rate of 4%.

关键词： Smart healthcare Wearable sensors Internet of Things (IoT) Deep Learning (DL) Progressive Residual Attention Based Holographic Convolutional neural Network (PRAHCNN) Cardiac disease risk

来源：评论

学校读者我要写书评

暂无评论

DehazeGS: 3D Gaussian Splatting for Multi-image Haze Removal

引用

IEEE signal processing LETTERS 2025年 32卷 736-740页

作者： Ma, Chenjun Zhao, Jieyu Chen, Jian Ningbo Univ Dept Comp Sci & Technol Ningbo 315211 Peoples R China

neural Radiance Fields (NeRF) have advanced 3D reconstruction by learning implicit representations of scenes from multi-view images, yet their effectiveness is limited in environments with scattering medium. Existing methods that incorporate scattering models into NeRF frameworks face issues with slow training speeds and high memory demands. This paper presents DehazeGS, a novel haze removal and reconstruction method based on 3D Gaussian Splatting (3DGS). Our approach integrates the Koschmieder scattering model into the 3DGS framework, enabling effective separation of objects and scattering medium. This method leverages a point-based representation to achieve high-quality scene reconstruction while significantly reducing computational and memory overhead. Experimental results on both synthetic and real datasets demonstrate that our method outperforms existing approaches in terms of dehazing quality and reconstruction performance, effectively synthesizing clear images from foggy scenes. Our findings suggest that integrating scattering models with 3DGS offers a promising solution for applications in adverse weather conditions.

关键词： Three-dimensional displays Scattering image color analysis neural radiance field Rendering (computer graphics) Covariance matrices Atmospheric modeling image reconstruction Computational modeling Point cloud compression 3D Gaussian splatting image dehazing neural radiance field novel view synthesis

来源：评论

学校读者我要写书评

暂无评论

Edge-guided and hierarchical aggregation network for robust medical image segmentation

引用

BIOMEDICAL signal processing AND CONTROL 2025年 101卷

作者： Tang, Yi Zhao, Di Pertsau, Dmitry Gourinovitch, Alevtina Kupryianava, Dziana Belarusian State Univ Informat & Radioelect Minsk 220013 BELARUS

Medical image segmentation with the convolutional neural networks (CNNs), significantly enhances clinical analysis and disease diagnosis. However, medical images inherently exhibit large intra-class variability, minimal inter-class differences, and substantial noise levels. Extracting robust contextual information and aggregating discriminative features for fine-grained segmentation remains a formidable task. Additionally, existing methods often struggle with producing accurate mask edges, leading to blurred boundaries and reduced segmentation precision. This paper introduces a novel Edge-guided and Hierarchical Aggregation Network (EHANet) which excels at capturing rich contextual information and preserving fine spatial details, addressing the critical issues of inaccurate mask edges and detail loss prevalent in current segmentation models. The Inter-layer Edge-aware Module (IEM) enhances edge prediction accuracy by fusing early encoder layers, ensuring precise edge delineation. The Efficient Fusion Attention Module (EFA) adaptively emphasizes critical spatial and channel features while filtering out redundancies, enhancing the model's perception and representation capabilities. The Adaptive Hierarchical Feature Aggregation Module (AHFA) module optimizes feature fusion within the decoder, maintaining essential information and improving reconstruction fidelity through hierarchical processing. Quantitative and qualitative experiments on four public datasets demonstrate the effectiveness of EHANet in achieving superior mIoU, mDice, and edge accuracy against eight other state-of-the-art segmentation methods, highlighting its robustness and precision in diverse clinical scenarios.

关键词： Medical image segmentation Efficient fusion attention Hierarchical feature aggregation Inter-layer edge-aware module

来源：评论

学校读者我要写书评

暂无评论

Gunet: a novel and efficient low-illumination palmprint image enhancement method

引用

signal image AND VIDEO processing 2024年第8-9期18卷 6093-6101页

作者： Zhou, Kaijun Lu, Duojie Liu, Guangnan Zhou, Xiancheng Qin, Yemei Hunan Univ Technol & Business Sch Intelligent Engn & Intelligent Mfg Yuelu Ave Changsha 410000 Hunan Peoples R China

Palmprint has a high application prospect due to the stability, uniqueness, difficulty of reproduction, easy acquisition and high user acceptance of its own texture characteristics. However, palmprint images acquired at long distances and in low light conditions lose a amount of significant palmprint texture features, which in turn will affect the accuracy of palmprint recognition. Combining traditional methods with deep learning methods, this paper proposes a low-illumination palmprint image enhancement method based on wavelet transform and GUNet (insertion of Integrated Attention Gate at the UNet jump connection). Firstly, the input image is decomposed by wavelet transformation to obtain the decomposition result of the original image, that is the high-frequency image and the low-frequency image. Secondly, the high-frequency image is enhanced by the GUNet neural network to enhance the palmprint texture, and low-frequency images use weighted averaging to smooth low-frequency information other than palm texture information. After that, the palmprint image with a clear palmprint texture is obtained by the wavelet inverse transformation. Numerous experiments are carried out on palmprint databases such as Idiap, CASIA, IITD and a laboratory self-collection. Experimental results show that the proposed approach has higher Peak signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM) and Visual Information Fidelity (VIF) indicators than some existing methods, which suggest palmprint image with low-illumination can be significantly enhanced.

关键词： GUNet Wavelet transform Long distance Low-illumination

来源：评论

学校读者我要写书评

暂无评论

Multifunctional adversarial examples: A novel mechanism for authenticatable privacy protection of images

引用

signal processing 2025年 230卷

作者： Li, Ming Wang, Si Henan Normal Univ Coll Comp & Informat Engn Xinxiang 453007 Peoples R China Key Lab Artificial Intelligence & Personalized Lea Xinxiang 453007 Henan Peoples R China

With the rapid development of network technology, more and more images containing personal identity characteristics are being released by users on open network platforms. However, these images are easily collected by malicious users, leading to problems such as privacy leakage, infringement, and tampering, thus harming users' legitimate interests. Recent studies have found that adversarial examples generated by adding tiny perturbations to an image can mislead image classifiers, causing incorrect classifications. Therefore significant privacy protection against deep neural networks is achieved while the visual quality remains indistinguishable to human eyes. However, these methods cannot protect the authenticity and integrity of the image simultaneously, failing to address infringement and tampering issues, which are also neglectable in the open network platforms. To solve this problem, we propose a novel authentication-enabled privacy protection method. The meaningful information used for authentication, instead of the meaningless perturbations, is embedded into the host image to generate adversarial examples, thereby achieving both authentication and privacy protection simultaneously. This scheme combines attention mechanisms with generative adversarial networks to adaptively select and weight features between different channels, achieving significant improvements in both aggressiveness and authentication capability. Experimental results show that our method outperforms recent similar methods in overall performance.

关键词： Privacy protection image authentication Adversarial examples Channel attention Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Improving time-frequency resolution in non-stationary signal analysis using a convolutional recurrent neural network

引用

signal image AND VIDEO processing 2024年第5期18卷 4797-4810页

作者： Krishna, B. Murali Satyanarayana, S. V. V. Satyanarayana, P. V. V. Suman, M. Venkata Sri Vasavi Engn Coll ECE Dept Tadepalligudem 534101 Andhra Pradesh India

In this paper, a convolutional recurrent neural network (ConvRNN) wavelet-based kernel is proposed to improve the time-frequency localization of non-stationary signals. Time-frequency distributions (TFDs) are used to localize the power spectral density components of a signal simultaneously over the time-frequency plane. For various non-stationary signals (NSS) with different features, poor time-frequency resolution is always an inherent problem in all developed TFDs. This is mainly due to inefficient segmentation methods and improper kernel adaptation. In the current work, the fraction Bezier-Bernstein polynomial function is applied to model the NSS, and points of inflection are used for signal segmentation. From the obtained segments, statistical time-frequency features are extracted and fed to a ConvRNN for better time-frequency localization. The ConvRNN employs a convolution computation between input signal features and the proposed Newton-Raphson gradient algorithm (NRGA)-based wavelet function. The optimization of the ConvRNN network is achieved by incorporating a hybrid method that combines the principles of normalized adaptive gradient descent and momentum-based optimization, with an additional normalization step to enhance convergence and stability. The ConvRNN weights are updated in both forward and backward directions (resilient propagation) until a better correlation is achieved between signal segments and the wavelet kernel. It is observed that the proposed ConvRNN NRGA wavelet improves the time-frequency localization when compared with standard TFDs and state-of-the-art methodologies. Furthermore, the proposed ConvRNN model is compared with other CNN and RNN architectures for better practical interpretations.

关键词： Bernstein polynomial Segmentation Wavelet Resolution Gradient

来源：评论

学校读者我要写书评

暂无评论

Exact Implementation of Closed-Form Liquid neural Networks With Arbitrary Precision

引用

IEEE signal processing LETTERS 2025年 32卷 921-925页

作者： Cantini, Clotilde Rolland-Piegue, Emilie Schmitter, Daniel Swiss Life Machine Learning & AI Grp CH-8022 Zurich Switzerland Ecole Ponts ParisTech F-77420 Champs sur Marne France

Continuous-time liquid neural networks constitute a novel class of machine learning models that emulate the dynamics of biological neurons and synapses using ordinary differential equations. Despite their promising applications in predicting spatiotemporal dynamics, the adoption of these models has been constrained by their reliance on computationally expensive numerical differential equation solvers or approximate solutions. In this work, we propose a redefinition of the network's core neuron that accommodates for multiple presynaptic connections. We then derive a closed-form solution and present an implementation through a computationally efficient recursive algorithm. Our solution is validated both at the level of individual neurons and within a deep neural network architecture. Experimental results on sequential modeling tasks with image, sensor or medical data, demonstrate improved performance compared to state-of-the-art numerical and approximate methods.

关键词： Neurons Synapses Biological neural networks Mathematical models Electrocardiography Training Liquids Data models Computational modeling Biological system modeling Closed-form continuous domain deep learning liquid time constant neural networks ordinary differential equation sampling

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：