- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Fast Image Super-resolution with Sparse Coding
摘要： In this paper, we introduce a novel fast image reconstruction method for super-resolution (SR) base on sparse coding. This method combine online dictionary learning and a fast sparse coding way, both of which can improve the efficiency of the reconstruction process and ensure the image visual quality. The new online optimization algorithm for dictionary learning based on stochastic approximations, which can drastically advance the learning speed, especially on millions of training samples. Meanwhile, we trained a neural network to speed up the reconstruction process, which based on iterative shrinkage-thresholding algorithm (ISTA), we called learned iterative shrinkage-thresholding algorithm (LISTA). It would produce best approximation sparse code with some fixed depth. We demonstrate that our approach can simultaneously improve the image fidelity and cost less computation.
关键词： super-resolution,fast image super-resolution,SRIR,sparse coding
更新于2025-09-04 15:30:14
-
[IEEE 2018 - 3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video (3DTV-CON) - Helsinki (2018.6.3-2018.6.5)] 2018 - 3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video (3DTV-CON) - A NOVEL DISPARITY-ASSISTED BLOCK MATCHING-BASED APPROACH FOR SUPER-RESOLUTION OF LIGHT FIELD IMAGES
摘要： Currently, available plenoptic imaging technology has limited resolution. That makes it challenging to use this technology in applications, where sharpness is essential, such as film industry. Previous attempts aimed at enhancing the spatial resolution of plenoptic light field (LF) images were based on block and patch matching inherited from classical image super-resolution, where multiple views were considered as separate frames. By contrast to these approaches, a novel super-resolution technique is proposed in this paper with a focus on exploiting estimated disparity information to reduce the matching area in the super-resolution process. We estimate the disparity information from the interpolated LR view point images (VPs). We denote our method as light field block matching super-resolution. We additionally combine our novel super-resolution method with directionally adaptive image interpolation from [1] to preserve sharpness of the high-resolution images. We prove a steady gain in the PSNR and SSIM quality of the super-resolved images for the resolution enhancement factor 8x8 as compared to the recent approaches and also to our previous work [2].
关键词： Light Field Image Super-Resolution,Block Matching,4D Imaging,Resolution Enhancement
更新于2025-09-04 15:30:14
-
Amorphous Quantum Nanomaterials
摘要： In quantum materials, macroscopic behavior is governed in nontrivial ways by quantum phenomena. This is usually achieved by exquisite control over atomic positions in crystalline solids. Here, it is demonstrated that the use of disordered glassy materials provides unique opportunities to tailor quantum material properties. By borrowing ideas from single-molecule spectroscopy, single delocalized π-electron dye systems are isolated in relatively rigid ultra-small (<10 nm diameter) amorphous silica nanoparticles. It is demonstrated that chemically tuning the local amorphous silica environment around the dye over a range of compositions enables exquisite control over dye quantum behavior, leading to efficient probes for photodynamic therapy (PDT) and stochastic optical reconstruction microscopy (STORM). The results suggest that efficient fine-tuning of light-induced quantum behavior mediated via effects like spin-orbit coupling can be effectively achieved by systematically varying averaged local environments in glassy amorphous materials as opposed to tailoring well-defined neighboring atomic lattice positions in crystalline solids. The resulting nanoprobes exhibit features proven to enable clinical translation.
关键词： photodynamic therapy (PDT),optical super-resolution microscopy,organic dyes,amorphous silica nanoparticles
更新于2025-09-04 15:30:14
-
[IEEE 2018 20th International Conference on Transparent Optical Networks (ICTON) - Bucharest (2018.7.1-2018.7.5)] 2018 20th International Conference on Transparent Optical Networks (ICTON) - Recent Advances in 3D Structured Illumination Microscopy
摘要： In structured illumination microscopy (SIM) the sample under investigation is illuminated using a structured illumination (SI) pattern. This SI pattern encodes high spatial frequencies of fine features within the sample, which usually are not transferred by the conventional three-dimensional (3D) optical transfer function (OTF) of the imaging system and fills the missing cone of frequencies in the OTF for better discrimination of the out-of- focus light. Thereby, SIM provides super-resolution (SR) performance beyond the diffraction limit and optical- sectioning (OS) capability with the use of data post-processing approaches. 3D structured patterns that include lateral and axial variations in the illumination have attracted more attention recently as they provide OS and SR enhancement in three dimensions. In this paper, we review recent implementations in generating a 3D SI pattern with tunable modulation frequency (independently of both the objective lens and the wavelength used) using a Fresnel biprism or a Wollaston prism. From the raw SIM images, final SIM images with simultaneous OS and SR are retrieved without artifacts due to coherent noise.
关键词： Fresnel biprism,Wollaston prism,structured illumination design,super-resolution,tunable frequency,three-dimensional microscopy
更新于2025-09-04 15:30:14
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Local Similarity Regularized Sparse Representation for Hyperspectral Image Super-Resolution
摘要： Recently, performance of hyperspectral image super-resolution (SR) has been significantly improved via sparse representation. However, most of these existing methods fail to consider the local geometrical structure of the sparse coefficients. To take this crucial issue into account, this paper proposes an effective method, which exploits the location related constraint about the sparse coefficients and incorporates their local similarity into the sparse coding process. Thus, the proposed method can preserve the properties of the aforementioned local geometrical structures. Based on the experimental results, the proposed method is demonstrated to be more effective than previous efforts in the task of hyperspectral image SR.
关键词： Local similarity,Sparse representation,Hyperspectral image,Super-resolution
更新于2025-09-04 15:30:14
-
Choosing the right label for single-molecule tracking in live bacteria: side-by-side comparison of photoactivatable fluorescent protein and Halo tag dyes
摘要： Visualizing and quantifying molecular motion and interactions inside living cells provides crucial insight into the mechanisms underlying cell function. This has been achieved by super-resolution localization microscopy and single-molecule tracking in conjunction with photoactivatable fluorescent proteins (PA-FPs). An alternative labelling approach relies on genetically-encoded protein tags with cell-permeable fluorescent ligands which are brighter and less prone to photobleaching than fluorescent proteins but require a laborious labelling process. Either labelling method is associated with significant advantages and disadvantages that should be taken into consideration depending on the microscopy experiment planned. Here, we describe an optimised procedure for labelling Halo-tagged proteins in live Escherichia coli cells. We provide a side-by-side comparison of Halo tag with different fluorescent ligands against the popular photoactivatable fluorescent protein PAmCherry. Using test proteins with different intracellular dynamics, we evaluated fluorescence intensity, background, photostability, and results from single-molecule localization and tracking experiments. Capitalising on the brightness and extended spectral range of fluorescent Halo ligands, we also demonstrate high-speed and dual-colour single-molecule tracking.
关键词： Halo tag,single-molecule tracking,photoactivatable fluorescent protein,Escherichia coli,fluorophores,super-resolution microscopy,DNA-binding proteins
更新于2025-09-04 15:30:14
-
Focusing and Super-Resolution with Partial Cloaking Based on Linear-Crossing Metamaterials
摘要： Manipulating the topological property of iso-frequency contour (IFC) will provide a unique control for the interaction between light and matter, such as increased rates of spontaneous emission and negative refraction. Different from the aforementioned topological transition that is realized by controlling the in-plane anisotropic electromagnetic parameters in a two-dimensional case, we theoretically propose and experimentally demonstrate another topological transition by tuning the vertical electromagnetic parameters of the anisotropic medium. Specially, the critical IFC in this transition corresponds to two intersecting lines and can be called linear-crossing metamaterial (LCMM). Based on the directional propagation and negative refraction in LCMM, we demonstrate the focusing and super-resolution with partial cloaking in transmission-line-based LCMMs. The realization of LCMM will open up another avenue for controlling light propagation and verifying unusual phenomena involving zero-index media and hyperbolic metamaterials.
关键词： partial cloaking,topological transition,super-resolution,linear-crossing metamaterial,focusing,iso-frequency contour
更新于2025-09-04 15:30:14
-
[IEEE 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE) - Nara, Japan (2018.10.9-2018.10.12)] 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE) - Super-Resolution Technique Utilizing A Non-linear Filter For Facial Images
摘要： Opportunities to process digital images on the Internet are increasing due to the popularization of recent innovations such as SNS and smartphones, resulting in the need for high-quality images. In this paper, we propose a facial-image processing method to obtain a beautiful facial image using a super-resolution technique. Specifically, a facial image is separated into a skin part and a non-skin part, and super-resolution processing is performed on the non-skin part, and facial correction processing is performed on the skin part. Experimental results show that high-quality, super-resolution images can be obtained.
关键词： Super-resolution,Facial correction
更新于2025-09-04 15:30:14
-
[IEEE 2018 International Conference on Advances in Computing, Communications and Informatics (ICACCI) - Bangalore, India (2018.9.19-2018.9.22)] 2018 International Conference on Advances in Computing, Communications and Informatics (ICACCI) - Enhanced Deep Image Super-Resolution
摘要： Recent advances in deep learning have facilitated new modalities for transforming the lower resolution image to higher resolution. The generated high resolution image must reconstruct the high frequency details of the image to generate a plausible result. To facilitate feature reuse for the task of super-resolution, we propose residual learning based convolutional neural network architecture. A pixel shuffle operation is performed in the upsampling procedure to mitigate the commonly encountered problem of artifacts in the predicted high resolution image. Our model makes use of a joint loss function consisting of pixel-wise loss and feature loss to learn the mapping from low resolution to its high resolution version. Additionally, our model has the ability to progressively increment to perform multi-scale super-resolution. An extensive experiment is performed to validate our model on the diverse ImageNet dataset. We show the effectiveness of our model through visual comparative assessment as well as quantitative comparative analysis with the state-of-the-art.
关键词： Residual block,Convolutional Neural Network,Image super-resolution
更新于2025-09-04 15:30:14
-
[IEEE 2018 IEEE International Conference on Consumer Electronics - Asia (ICCE-Asia) - JeJu, Korea (South) (2018.6.24-2018.6.26)] 2018 IEEE International Conference on Consumer Electronics - Asia (ICCE-Asia) - Accurate License Plate Recognition and Super-Resolution Using a Generative Adversarial Networks on Traffic Surveillance Video
摘要： Automatic License Plate Recognition (ALPR) is one of the most important methods of intelligent traffic surveillance applications. Some existing ALPR systems are developed for near-frontal plate images in a single lane. However, most surveillance cameras have a challenging environment: small size object, poor resolution and blurred image. We propose a new method that can be applied in the ALPR challenged environments by using super-resolution (SR) module based on Generative Adversarial Networks (GAN). We also used the state-of-the-art and real-time object detection method, You Only Look Once (YOLO), for license plate detection and character recognition. We collected a challenging dataset at low resolution and small object less than 60*60 size and evaluate our approach on it. The achieved mean accuracy of recognition of license plate is above 2% better than other methods in our dataset. Our implementation demonstrate the superiority over the state-of-the-art.
关键词： Visual Surveillance,Generative Adversarial Networks,License Plate Recognition,Super-Resolution
更新于2025-09-04 15:30:14