With the continuous advancement of technology, array camera systems, serving as a collaborative system of multiple cameras, exhibit broad prospects in areas such as surveillance, imageprocessing. and computer vision....
详细信息
作者：
Tong, LongyongZhou, HangSheil, BrianChongqing Univ
Coll Civil Engn Key Lab New Technol Construct Cities Mt Areas Chongqing 400045 Peoples R China Univ Cambridge
Laing ORourke Ctr Construct Engn & Technol Dept Engn Construct Engn Cambridge CB2 1PZ England
This paper presents a new approach for measuring large deformations in geotechnical experiments employing digitalimage correlation (DIC) or particle image velocity (PIV) techniques. The proposed method is based on th...
详细信息
This paper presents a new approach for measuring large deformations in geotechnical experiments employing digitalimage correlation (DIC) or particle image velocity (PIV) techniques. The proposed method is based on the Eulerian analysis scheme, allowing for the application of multicore central processing.unit (CPU)-based parallel computing to expedite the processing.of experimental images. The displacement increments obtained through DIC analysis on the Eulerian mesh nodes (subset centers) are then mapped onto tracer particles (TPs), which are assigned by users to track material movement. Finally, accumulated displacements and strains are determined on these TPs. Two example applications are presented to showcase the capabilities of the proposed method: a centrifuge half model test of flat circular footing penetrating sand overlying clay and a full transparent soil model test (TMST) of conical pile penetration. A comparison with other standard first-order deformation algorithms and the Lagrangian analysis scheme demonstrates that the presented method offers comparable precision but significantly faster computation speed, with an improvement of over six times when processing.a considerable number of (e.g. over 20) images. This enhanced computational speed can greatly reduce the time required for image post-processing. The proposed method is particularly suitable for large deformation experiments that involve the analysis of numerous images and require high precision.
Photograph processing.is a discipline of laptop science that concentrates on manipulating digital photos and using algorithms to recognize and beautify the photo's quality. Contour coding is a powerful approach us...
Photograph processing.is a discipline of laptop science that concentrates on manipulating digital photos and using algorithms to recognize and beautify the photo's quality. Contour coding is a powerful approach used in virtual picture processing.that allows one to increase, shop, and examine virtual snapshots from a combination of strategies like aspect detection, morphological processing. and segmentation. Contour coding aims to reduce complex virtual pix to a fixed of fundamental symbolic features that may be created, stored, and manipulated without loss in quality or element. Applications of contour coding are specially used for item reputation, segmentation, spatial pattern reputation, and scientific photograph processing. This newsletter offers the theory and fundamentals of contour coding and explores its software in actual-world virtual picture processing.scenarios. We speak about the significance of contour coding and present a short evaluation of existing algorithms and improvement techniques used to achieve virtual photograph processing.desires. We additionally gift the advances and challenges of contour coding and how it could form the future of digital picture processing.
digitalimageprocessing.technology has gone through rapid development and is extensively applied in daily life and production, with the rapid development of modern information technology. It plays an inestimable role...
详细信息
digitalimageprocessing.technology has gone through rapid development and is extensively applied in daily life and production, with the rapid development of modern information technology. It plays an inestimable role in remote sensing, medicine, recognition and other fields. This paper briefly introduces the basic concept of digitalimageprocessing.summarizes and analyses the commonly used digitalimageprocessing.technology and the latest scientific research achievements from four aspects, and puts forward the future development direction of digitalimageprocessing. In the future, it will pay more attention to artificial intelligence algorithms and achieve better processing.results by optimizing the logical *** using the simplified image algorithm, the application scope of digitalimageprocessing.will gradually expand, and will develop in the direction of miniaturization, intelligence, and convenience.
digital imaging is omnipresent nowadays due to the expansion of the mobile device industry and its use cases. However, the diversity of sensors and hardware limitations often lead to images not meeting desired quality...
详细信息
digital imaging is omnipresent nowadays due to the expansion of the mobile device industry and its use cases. However, the diversity of sensors and hardware limitations often lead to images not meeting desired quality standards. To counteract this, significant research has been invested in image de-noising and sharpening to enhance the quality of captured images, making image sharpening an integral step in professional imageprocessing. Nonetheless, certain challenges persist, including the frequent occurrence of over-enhancement in parts of the image, giving rise to artefacts such as ''jaggies'' or jagged edges, and a ''halo'' effect. Recognizing the successful outcomes achieved when dilated or extended filters are employed in other imageprocessing.tasks, like edge detection, our study is set to investigate a variety of sharpening algorithms using these filter extensions. A comprehensive evaluation was conducted on synthetic and natural images using a variety of image quality metrics, like PSNR or BRISQUE. In most cases, dilated kernels outperformed extended kernels, suggesting their potential superiority in image sharpening tasks while minimizing the introduction of undesired artefacts. However, the choice between dilated and extended kernels would ultimately depend on the specific application and algorithm used.
Traditional von Neumann architecture, characterized by separate memory and processing.units connected by a limited-capacity memory bus, faces challenges in handling tasks with high energy efficiency requirements. In c...
详细信息
Traditional von Neumann architecture, characterized by separate memory and processing.units connected by a limited-capacity memory bus, faces challenges in handling tasks with high energy efficiency requirements. In contrast, the compute-in-memory (CIM) architecture offers a promising alternative, facilitating high parallelism in data processing.while integrating storage functions, thereby significantly reducing memory access frequency and power consumption. This study presents a fully digital CIM macro featuring a novel self-write-back 12T cell. This bitcell is capable of performing Boolean logic operations and autonomously writing back results into the insitu cell, thereby significantly improving energy *** can be used for binary logical operations between each pixel of two images without requiring additional storage area. A bidirectional read/write architectural design enables matrix transposition. This can achieve image ***, this novel 12T cell also offers the option to choose not to write back the results of logical operations, thereby providing the flexibility and configurability for imageprocessing. Combined with an adder tree, it enables multiply-and-accumulate (MAC) operation for convolutional neural networks(CNNs). This can be used for feature extraction. We propose an 18T full adder structure with lower power-delay product than current state-of-the-art full adders, which is advantageous for improving the synthesis performance of the adder *** CIM macro supports a 4-bit x 4-bit MAC operation. The proposed CIM macro, designed and simulated using a 28 nm CMOS process with a 16 Kb static random access memory (SRAM), demonstrates promising results. At VDD = 0.9 V, the logic operation energy consumption is 1.35 fJ/bit with an energy efficiency of 740TOPS/W, and MAC operation energy consumption is 15.91 fJ/bit with an energy efficiency of 62.85 TOPS/W.
Due to the increasing demand for artificial intelligence technology in today's society, the entire industrial production system is undergoing a transformative process related to automation, reliability, and robust...
详细信息
Due to the increasing demand for artificial intelligence technology in today's society, the entire industrial production system is undergoing a transformative process related to automation, reliability, and robustness, seeking higher productivity and product competitiveness. Additionally, many hardware platforms are unable to deploy complex algorithms due to limited resources. To address these challenges, this paper proposes a computationally efficient lightweight convolutional neural network called Brightness Improved by Light-DehazeNet, which removes the impact of fog and haze to reconstruct clear images. Additionally, we introduce an efficient hardware accelerator architecture based on this network for deployment on low-resource platforms. Furthermore, we present a brightness visibility restoration method to prevent brightness loss in dehazed images. To evaluate the performance of our method, extensive experiments were conducted, comparing it with various traditional and deep learning-based methods, including images with artificial synthesis and natural blur. The experimental results demonstrate that our proposed method excels in dehazing ability, outperforming other methods in comprehensive comparisons. Moreover, it achieves rapid processing.speeds, with a maximum frame rate of 105 frames per second, meeting the requirements of real-time processing.
digital pictures are modified during the process of image enhancement to provide outcomes that are more suited for display or additional image analysis. Up to now, many achievements have been made in the field of imag...
详细信息
digital pictures are modified during the process of image enhancement to provide outcomes that are more suited for display or additional image analysis. Up to now, many achievements have been made in the field of image enhancement processing.image enhancement algorithms is consisted of deblurring, filtering, and contrast methods. And Gaussian filtering is the most widely used image enhancement algorithm among all. It is a linear filtering methods which serves to image smoothing and noise -removing But at the same time, it may lose details or produce halos when decomposing images, and it will not work well when processing.images affected by strong light or insufficient illumination. Nowadays, deep learning algorithms are applied in the image enhancement too. Neural networks like AlexNet and many other convolutional neural networks. The application of the CNN helps to solve the problem of limited accuracy but at the same time it cost more time for calculation. Also, the gray value of the pixel is not considered. To address the issue of image enhancement under complex lighting environments, we utilize an image enhancement approach based on bilateral filtering to achieve effective image enhancement in too -dark or too -bright lighting and also remove the problem of blurring edge caused by Gaussian filtering. Bilateral filtering could remove the noise and remain the information of the edges at the same time. Due to its own mathematical properties, it has the advantage to process the color image. Also, bilateral filtering could be beneficial to the process of edge -based surveillance videos. The efficiency of the surveillance videos has always been an issue. At the most of the time, the video could be vague and the problem of over exposure is more frequent than the tranquil image. For the edge -based surveillance video, the most important thing is speed and the clarity. And bilateral filtering offers one of the solution to the enhancement of the video and make it more cl
Tuberculosis affects various tissues, including the lungs, kidneys, and brain. According to the medical report published by the World Health Organization (WHO) in 2020, approximately ten million people have been infec...
详细信息
Tuberculosis affects various tissues, including the lungs, kidneys, and brain. According to the medical report published by the World Health Organization (WHO) in 2020, approximately ten million people have been infected with tuberculosis. U-NET, a preferred method for detecting tuberculosis-like cases, is a convolutional neural network developed for segmentation in biomedical imageprocessing. The proposed RNGU-NET architecture is a new segmentation technique combining the ResNet, Non-Local Block, and Gate Attention Block architectures. In the RNGU-NET design, the encoder phase is strengthened with ResNet, and the decoder phase incorporates the Gate Attention Block. The key innovation lies in the proposed Local Non-Local Block architecture, overcoming the bottleneck issue in U-Net models. In this study, the effectiveness of the proposed model in tuberculosis segmentation is compared to the U-NET, U-NET+ResNet, and RNGU-NET algorithms using the Shenzhen dataset. According to the results, the RNGU-NET architecture achieves the highest accuracy rate of 98.56%, Dice coefficient of 97.21%, and Jaccard index of 96.87% in tuberculosis segmentation. Conversely, the U-NET model exhibits the lowest accuracy and Jaccard index scores, while U-NET+ResNet has the poorest Dice coefficient. These findings underscore the success of the proposed RNGU-NET method in tuberculosis segmentation.
Earth image decomposition plays an important role in understanding and interpreting high-resolution SAR images. Abundant information provided by SAR images allows different models of earth materials to be presented di...
详细信息
Earth image decomposition plays an important role in understanding and interpreting high-resolution SAR images. Abundant information provided by SAR images allows different models of earth materials to be presented directly using spatial and textural structures from machine learning methods. Since weather diagnosis and forecasting are mostly done through aerial and satellite images in meteorology, imageprocessing.is widely used in this science and greatly increases the accuracy and speed of forecasting weather and storms. Aerial and satellite images play an important role in predicting residential areas and structures, and processing.these images plays a significant role in the accuracy and speed of prediction. In this thesis, a general review was done of the types of algorithms available in SAR and PolSAR imageprocessing. With the superpixel-based LCDFL (locality-constraint discriminant feature learning) unsupervised learning method, image features are selected for classification. Finally, the images were classified by training the features using the support vector machine. The proposed method is simulated on POLSAR images with MATLAB software. Among the advantages of the algorithm are that the recovery speed is better on a home computer, and the results show that using the unsupervised learning method gives an accuracy of image classification of around 95% compared to similar performance methods. It has been better.
暂无评论