Underwater images are challenging for correspondence search algorithms, which are traditionally designed based on images captured in air and under uniform illumination. In water however, medium interactions have a muc...
详细信息
ISBN:
(纸本)9783031064333;9783031064326
Underwater images are challenging for correspondence search algorithms, which are traditionally designed based on images captured in air and under uniform illumination. In water however, medium interactions have a much higher impact on the light propagation. Absorption and scattering cause wavelength- and distance-dependent color distortion, blurring and contrast reductions. For deeper or turbid waters, artificial illumination is required that usually moves rigidly with the camera and thus increases the appearance differences of the same seafloor spot in different images. Correspondence search, e.g. using image features, is however a core task in underwater visual navigation employed in seafloor surveys and is also required for 3D reconstruction, image retrieval and object detection. For underwater images, it has to be robust against the challenging imaging conditions to avoid decreased accuracy or even failure of computer vision algorithms. However, explicitly taking underwater nuisances into account during the feature extraction and matching process is challenging. On the other hand, learned feature extraction models achieved high performance in many in-air problems in recent years. Hence we investigate, how such a learned robust feature model, D2Net, can be applied to the underwater environment and particularly look into the issue of cross domain transfer learning as a strategy to deal with the lack of annotated underwater training data.
Digital images are affected by a variety of noise and one well-known type is impulsive noise. In order to reduce or eliminate noise, many image-denoising algorithms have been created, with varying benefits and limitat...
详细信息
ISBN:
(数字)9798350369748
ISBN:
(纸本)9798350369755
Digital images are affected by a variety of noise and one well-known type is impulsive noise. In order to reduce or eliminate noise, many image-denoising algorithms have been created, with varying benefits and limitations. To deal with impulsive and spurious noise in colour images, this study does a thorough examination with an emphasis on the median filter and its various variants. By means of thorough experimentation, the researchers examine the relative performance of different denoising algorithms using Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR) as standards. The results of the study demonstrate the usefulness of the Vector Median filter, especially in situations with high-density impulsive noise. The Vector Median filter is particularly effective for real-time imageprocessing applications since it performs better and requires less processing time. Furthermore, the Modified Median filter, with its high PSNR and low MSE values, shows potential as a low-density noise solution. This study offers insightful information about image eliminating techniques aimed at reducing impulsive noise in colour images. The study advances the field of imageprocessing by utilising creative methodologies and performance measurements. This has significance for other sectors that depend on accurate image analysis. The study also sets the basis for next investigations focused on improving and expanding the range of denoising algorithms to handle a greater variety of noise kinds and intensities.
Snapshot multispectral imaging systems typically capture multispectral images in a single shot by covering the sensor with a multispectral filter array (MSFA). A demosaicking algorithm is generally required for such s...
详细信息
ISBN:
(数字)9798331515669
ISBN:
(纸本)9798331515676
Snapshot multispectral imaging systems typically capture multispectral images in a single shot by covering the sensor with a multispectral filter array (MSFA). A demosaicking algorithm is generally required for such systems to reconstruct the full-resolution multispectral images. A two-stage demosaicking method is proposed in this paper. In the first stage, the image is progressively interpolated by Neville filters. In the second stage, the directional interpolation with a novel weighting function is performed to enhance the quality of the pre-interpolated image. Experimental results demonstrate that the proposed approach achieves superior performance in both subjective and objective assessments.
A lightweight neural network-based approach to two-person interaction classification in image sequences, based on human skeletons detected in sparse video frames, is proposed. The idea is to use an ensemble of pose cl...
详细信息
ISBN:
(纸本)9788396242396
A lightweight neural network-based approach to two-person interaction classification in image sequences, based on human skeletons detected in sparse video frames, is proposed. The idea is to use an ensemble of pose classifiers ("experts"), where every expert is trained on different time-indexed snapshots of an interaction. Thus, the expertise of "weak" classifiers is distributed over the time duration of an interaction. The overall classification result is a weighted combination of all the pose experts. Important element of proposed solution is the refinement of skeleton data, based on a merging-of-joints procedure. This allows the generation of reliable features being passed to the artificial neural network. This is the key to our lightweight solution, as ANN resources, needed for feature space transformation, can be significantly limited. Our network model was trained and tested on the interaction subset of the well-known NTU RGB+D dataset, although only 2D skeleton information is used, typical in video analysis. The test results show comparable performance of our method with some of the best so far reported STMand CNN-based classifiers for this dataset, when they process sparse frame sequences, like we did. The recently proposed multistream Graph CNNs have shown superior results but only when processing dense frame sequences. Considering the dominating processing time and resources needed for skeleton estimation in every frame of the sequence, the key to real-time interaction recognition is to limit the number of processed frames.
China has seen an unheard-of surge in interest in deep-learning methods for image restoration in recent years. Most of these strategies draw inspiration from the established variational technique and related optimizat...
详细信息
ISBN:
(数字)9798350386639
ISBN:
(纸本)9798350386646
China has seen an unheard-of surge in interest in deep-learning methods for image restoration in recent years. Most of these strategies draw inspiration from the established variational technique and related optimization methods for the picture reconstruction inverse issue. While using learnable components to create organized deep neural networks and using copious amounts of observation data to train the networks for the particular reconstruction objectives, these techniques resemble the iterative strategies of ordinary optimization algorithms. In many cases, they have proven to have far better empirical performance than the conventional approaches, and they also demand a lot lower computing cost. For various common networks in this subject, this research offers the specifics of the derivations, the network topologies, and the training protocol. The research therefore focuses on imageprocessingalgorithms based on variational and deep learning models.
The proceedings contain 87 papers. The topics discussed include: Auto-FP: an experimental study of automated feature preprocessing for tabular data;Data-CASE: grounding data regulations for compliant data processing s...
ISBN:
(纸本)9783893180943
The proceedings contain 87 papers. The topics discussed include: Auto-FP: an experimental study of automated feature preprocessing for tabular data;Data-CASE: grounding data regulations for compliant data processingsystems;data coverage for detecting representation bias in image datasets: a crowdsourcing approach;balancing utility and fairness in submodular maximization;stateful entities: object-oriented cloud applications as distributed dataflows;learning over sets for databases;a new PET for data collection via forms with data minimization, full accuracy and informed consent;adaptive compression for databases;analysis of open government datasets from a data design and integration perspective;fine-grained geo-obfuscation to protect workers’ location privacy in time-sensitive spatial crowdsourcing;and a framework to evaluate early time-series classification algorithms.
The proceedings contain 87 papers. The topics discussed include: Auto-FP: an experimental study of automated feature preprocessing for tabular data;Data-CASE: grounding data regulations for compliant data processing s...
ISBN:
(纸本)9783893180943
The proceedings contain 87 papers. The topics discussed include: Auto-FP: an experimental study of automated feature preprocessing for tabular data;Data-CASE: grounding data regulations for compliant data processingsystems;data coverage for detecting representation bias in image datasets: a crowdsourcing approach;balancing utility and fairness in submodular maximization;stateful entities: object-oriented cloud applications as distributed dataflows;learning over sets for databases;a new PET for data collection via forms with data minimization, full accuracy and informed consent;adaptive compression for databases;analysis of open government datasets from a data design and integration perspective;fine-grained geo-obfuscation to protect workers’ location privacy in time-sensitive spatial crowdsourcing;and a framework to evaluate early time-series classification algorithms.
Vedic Multiplier is a key tool in rapidly growing technology especially in the immense domain of imageprocessing, Digital Signal processing, real-time signal. Multipliers are important block in digital systems and pl...
详细信息
Data exchanges can be significant in the Deep Neural Network (DNN) algorithms. The interconnection between computing resources can quickly have a substantial impact on the overall performance of the architecture and i...
详细信息
ISBN:
(纸本)9783031299698;9783031299704
Data exchanges can be significant in the Deep Neural Network (DNN) algorithms. The interconnection between computing resources can quickly have a substantial impact on the overall performance of the architecture and its energy efficiency. Similarly, access to the different memories of the system, with potentially high data sharing, is a critical point. To overcome these problems, in this paper, we propose a new interconnect network, called AINoC, for future DNN accelerators, which require more flexibility and less power consumption to facilitate their integration into artificial intelligence (AI) edge systems. AINoC is based on (1) parallel routing that ensures communication/computation overlap to improve performance and (2) data reuse (filters, image inputs, and partial sums) to reduce multiple memory accesses. In the experiment section, AINoC can speedup LeNet5 convolution layers by up to 71.74x in latency performance w.r.t. a RISC-V-based CPU and also speedup MobileNetV2 convolution layers by up to 2.35x in latency performance w.r.t. a dataflow architecture featuring row-stationary execution style. AINoC provides any-to-any data exchange with wide interfaces (up to 51.2 GB/s) to support long bursts (up to 384 flits/cycle with packed data, i.e., 3 * 8-bit data in a 32-bit wide datapath) while executing LeNet5 and MobileNetV2. AINoC supports flexible communication with many multicast/broadcast requests with non-blocking transfers. Parallel communication in AINoC can provide up to 128x more throughput (flits/cycle) and bandwidth (GB/s), using parallel routing with respect to single-path routing while executing convolution layers of LeNet5 and MobiletNetV2.
Noise poses a maj or challenge to imageprocessing, making accurate analysis and interpretation more difficult. Anisotropic diffusion algorithms specifically tailored for noisy images across several domains are examin...
详细信息
ISBN:
(数字)9798350364828
ISBN:
(纸本)9798350364835
Noise poses a maj or challenge to imageprocessing, making accurate analysis and interpretation more difficult. Anisotropic diffusion algorithms specifically tailored for noisy images across several domains are examined in this paper. The efficiency of anisotropic diffusion in lowering noise in various image datasets is evaluated in-depth in this study. Based on comprehensive study and experimentation, this work presents real proof of the efficacy of this method in decreasing noise while preserving significant image properties in multiple domains. In comparison to the MP and MPM models, the experimental findings show that the suggested model performs quite well.
暂无评论