Compressive sensing (CS) for a full-sized image requires a large sensing matrix, consuming significant storage space and computational resources. Block-based CS addresses this but allocates the same number of samples ...
详细信息
ISBN:
(数字)9798331541460
ISBN:
(纸本)9798331541477
Compressive sensing (CS) for a full-sized image requires a large sensing matrix, consuming significant storage space and computational resources. Block-based CS addresses this but allocates the same number of samples to each block, ignoring the sparsity differences among various blocks. Additionally, the original image's sparsity is unknown in practical systems. Therefore, it is necessary to accurately estimate the sparsity from some initial sampling information. This paper proposes an adaptive rate block CS model based on the difference of the results of two reconstruction algorithms. The image is initially sampled at the sampling end. Then a high-accuracy algorithm and a low-accuracy algorithm are used to reconstruct initial sampled signals respectively. The sparsity of each image block is estimated by the difference between the two initial reconstructed signals, allowing the appropriate allocation of the number of samples for a block. The main computations of this scheme are concentrated at the reconstruction end, which can effectively save resources at the sampling end. Experimental results show that the proposed scheme can effectively improve reconstruction quality while reducing the sampling rate.
Photothermal power generation is a promising technique for converting solar radiation into electricity with high efficiency and stability. However, the performance and maintenance of photothermal power plants depend o...
详细信息
ISBN:
(数字)9798350366556
ISBN:
(纸本)9798350366563
Photothermal power generation is a promising technique for converting solar radiation into electricity with high efficiency and stability. However, the performance and maintenance of photothermal power plants depend on the cleanliness and reflectivity of the heliostats. This paper introduces an innovative approach to addressing the challenges of dirt detection and segmentation on heliostats. Leveraging the capabilities of deep learning, we propose the Multi-Scale Heliostat Dirt Segmentation and Classification (MSHDSC) framework, integrating a novel multi-scale feature fusion module (MSFFM) with an enhanced DeepLabV3+ network. This framework effectively segments small dirt areas on heliostat images, facilitating precise cleaning strategies. A unique aspect of our work is the introduction of an unsupervised clustering algorithm post-segmentation, which categorizes dirt based on color and texture, assigning a severity score to each category. This categorization assists in determining the cleaning complexity and prioritizing maintenance efforts. Experimental results show that our method outperforms several state-of-the-art image segmentation models in terms of accuracy and efficiency and provides useful information for targeted and prioritized cleaning of heliostats by robots or drones.
Machine vision and computer imageprocessing technologies are widely used in the metallurgical industry, especially in recognizing and analyzing defects in glass. High surfaces of planer surface and quality in the gla...
详细信息
ISBN:
(数字)9798331527662
ISBN:
(纸本)9798331527679
Machine vision and computer imageprocessing technologies are widely used in the metallurgical industry, especially in recognizing and analyzing defects in glass. High surfaces of planer surface and quality in the glass and metal manufacturing sector primarily require automated and performance visual detection, and inspection systems and algorithms are constantly improving. This article, therefore, seeks to comprehensively recognize and analyze glass subsurface defects through the integration of imageprocessing technologies and machine vision. According to the algorithm's properties and the image features typified in the process, the defects of identified glass components are primarily classified based on determining the cause and the optical features of characteristics exhibited. Given the nature of varied and detected effects determined mainly by the classifiers, an improved accuracy level was notable, and a reduction of the computational complexity was evident following the use of machine vision and imageprocessing technologies.
Indoor visible light positioning (VLP) systems based on received signal strength (RSS) fingerprint can provide high-precision indoor positioning with low complexity. However, laborious offline field measurements are r...
Indoor visible light positioning (VLP) systems based on received signal strength (RSS) fingerprint can provide high-precision indoor positioning with low complexity. However, laborious offline field measurements are required to collect RSS fingerprints, and the subsequent fingerprint matching process also needs a large computation overhead. In order to address the above problems, in this paper, we propose an indoor VLP method based on sparse fingerprints using extended Min-Max (E-Min-Max) and weighted k-nearest neighbor (WKNN) algorithms. In the proposed VLP method, bicubic interpolation is used during the offline phase to directly estimate the RSS values of virtual reference nodes. In the online phase, a coarse estimate of the position of target node (TN) is firstly obtained by using the E-Min-Max algorithm. Then, based on the coarse estimate of the TN’s position, the WKNN algorithm is used to determine the TN’s accurate position. Simulation results show that compared to the traditional WKNN algorithm, the proposed VLP method has much higher fingerprint matching efficiency while ensuring high positioning accuracy.
Aurora phenomenon is a projection of solar-terrestrial interaction onto the polar upper atmosphere. Systematically observing the auroral morphology and its evolution helps humans study solar activity and space weather...
详细信息
ISBN:
(数字)9798350349115
ISBN:
(纸本)9798350349122
Aurora phenomenon is a projection of solar-terrestrial interaction onto the polar upper atmosphere. Systematically observing the auroral morphology and its evolution helps humans study solar activity and space weather. To enhance more details of auroral morphology and motion, which is essential for cross-site collaborative studies and identification of small-scale structures, we employ a super-resolution (SR) approach to improve the resolution of all-sky auroral images. Considering that similar auroral structures and motion patterns may appear in both neighboring frames and the current image, we introduce spatio-temporal continuity into the image super-resolution reconstruction algorithm and explore the nonlocal similarity between neighboring frames to obtain complete and real high-resolution auroral structures. An adaptive all-sky field mapping strategy is proposed to change its neighborhood relationship to obtain a low-rank relationship in all-sky field. A low-rank based super-resolution approach is applied to global images containing spatial-temporal continuity. We experimentally show that the proposed method improves the details of reconstructed high-resolution images and outperforms other methods for ASI auroral images.
Underwater image restoration is conducive to better underwater resource detection and information effective ***,the light in the complex water body diffusely reflective and the selection absorption of different band l...
详细信息
ISBN:
(数字)9789887581536
ISBN:
(纸本)9781665482561
Underwater image restoration is conducive to better underwater resource detection and information effective ***,the light in the complex water body diffusely reflective and the selection absorption of different band light results in blurring and color distortion of underwater ***,we propose a convolutional Dark Channel Prior(DCP)underwater image recovery algorithm to enhance underwater *** can do the pre-processing work for the subsequent YOLOv5 object *** enhancement algorithm first performs Commission International Eclairage Lab(CIELAB)equalization of underwater images for color distortion ***,the underwater image formation parameters are estimated by the minimum convolution region ***,Contrast Limited Adaptive Histogram Equalization(CLAHE) is performed to obtain an enhanced underwater ***,the enhanced underwater image is input to the YOLOv5 model for *** results show that the proposed method outperforms state-of-art algorithms in terms of image recovery effect,evaluation quality and detection accuracy.
The proceedings contain 51 papers. The special focus in this conference is on Advances in Information Communication Technology and Computing. The topics include: Internet of Things and Retail Performance in an Emergin...
ISBN:
(纸本)9789811998874
The proceedings contain 51 papers. The special focus in this conference is on Advances in Information Communication Technology and Computing. The topics include: Internet of Things and Retail Performance in an Emerging Market: A Qualitative Analysis;the Impact of Social Media on Consumer Purchase Behaviour;the Contribution of the Internet of Things to Enhance the Brands of Small and Medium-Sized Enterprises in Iraq;trust Chain for Managing Trust in Blockchain-Associated IoT-Enabled Supply-Chains;supply Chain Management Using Blockchain Security Enhancement;a Novel image Encryption Algorithm Using Logistic and Henon Map;an Analysis of Data Sparsity Resolution algorithms Used in Recommender systems;recent Trends for Practicing Machine Learning in Brain Tumors: A Survey;voice-Based Intelligent Virtual Assistant;The Impact of IT Capabilities on Competitive Advantage;Identification of Autism Spectrum Disorder from Functional MRI Using Deep Learning;conversion of Sign Language to Text and Audio Using Deep Learning Techniques;security Attacks and Key Challenges in Blockchain Technology: A Survey;advancing from Manual to Automatic Telecast of News for Deaf;study of Nutrition-Based Recommender System for Diabetes and Cardiovascular Patients Based on Various Machine Learning Techniques: A Systematic Review;comparing Fish Finding Techniques using Satellite and Indigenous Data based on Different Machine Learning algorithms;implementation of E-Governance Framework for Rural Areas of India;opinion Summarization from Online Mobile Phone Reviews Using Feature Based Association Rule Mining;unconventional to Automated Attendance Marking Using imageprocessing;synthetic Animations Based Dictionary for Indian Sign Language;start of Telemedicine in Uzbekistan. Technological Availability;review of Feature Extraction Techniques for Fake News Detection.
Total focusing method (TFM), as the most effective post-processing technique for ultrasonic phased array imaging currently, can achieve focusing at any point within the detection area, possessing advantages of high re...
详细信息
ISBN:
(数字)9798331516550
ISBN:
(纸本)9798331516567
Total focusing method (TFM), as the most effective post-processing technique for ultrasonic phased array imaging currently, can achieve focusing at any point within the detection area, possessing advantages of high resolution and signal-to-noise ratio that traditional phased array imaging methods lack. However, Total Focusing Method (TFM) imaging requires delay-and-sum processing of full matrix capture (FMC) data. If the number of elements involved in signal acquisition is N, the full matrix data will contain N A-scan signals, each containing thousands of sampling *** will result in long computation time and make it difficult to achieve real-time imaging. This paper proposes a fast total focusing method that combines elliptical arc scanning algorithm and OpenCL hardware acceleration, and deploys it to a self-developed ultrasound phased array imaging system. Results show that the fast total focusing method can generate a 512 × 512 pixel image with 32 elements in an average time of only 64 milliseconds, far surpassing the time required by conventional total focusing algorithms, providing a feasible approach for real-time high-precision non-destructive testing online.
This article proposes a distributed video AI super-resolution reconstruction solution based on MapReduce to address the issue of slow video super-resolution reconstruction in a single-machine environment. The solution...
详细信息
This paper presents a real-time embedded thermal imaging system architecture for compact, energy-efficient, high-quality imaging utilizing heterogeneous system-on-chip (SoC) and uncooled infrared focal plane arrays (I...
详细信息
ISBN:
(数字)9798350387957
ISBN:
(纸本)9798350387964
This paper presents a real-time embedded thermal imaging system architecture for compact, energy-efficient, high-quality imaging utilizing heterogeneous system-on-chip (SoC) and uncooled infrared focal plane arrays (IRFPAs). Unlike previous systems that organized separate devices for complex imageprocessing, our system provides integrated imageprocessing support for robust sensor-to-surveillance. The imageprocessing organizes two algorithm stacks: a non-uniformity correction stack to mitigate the distinctive noise vulnerabilities of uncooled IRFPAs, and an image enhancement stack including contrast enhancement and temporal noise filters. We optimized these algorithms for domain-specific factors, including asymmetric multiprocessing (AMP), cache organization, single instruction multiple data (SIMD) instructions, and very long instruction word (VLIW) architectures. The implementation on the TI TDA3x SoC demonstrates that our system can process 640×480, 60 frames per second (FPS) videos at a peak core load of 57.5% while consuming power less than 2.2 W for the entire system, denoting the possibility of processing the 1280×1024, 30 FPS videos from the cutting-edge uncooled IRFPAs. Additionally, our system improves power efficiency by 9.42% and 9.96% at 30 and 60 FPS, respectively, compared to the state-of-the-art when executing similar imageprocessingalgorithms.
暂无评论