This study proposes a way to detect vitamin deficiency by combining machine learning and imageprocessing. Computer vision enables the system to recognise visual symptoms of specific vitamin deficiencies. The recommen...
详细信息
ISBN:
(数字)9798331539948
ISBN:
(纸本)9798331539955
This study proposes a way to detect vitamin deficiency by combining machine learning and imageprocessing. Computer vision enables the system to recognise visual symptoms of specific vitamin deficiencies. The recommended approach is that the entire procedure can be subdivided into specific key steps, which are initiated from image acquisition, followed by the image preprocessing steps used to enhance their quality. It catches the confusing patterns that are directed toward various abnormalities through a pretrained Convolutional Neural Network (CNN) model. Finally, with such patterns at hand, the categorisation takes place, which in turn helps to identify specific *** extensive experimentation across a diverse dataset, the system demonstrates remarkable accuracy in detecting deficiency. Its non-invasive nature permits early screening. This proves its potential for widespread implementation and directions for future enhancement, such as dataset expansion and exploration of other advanced architectures apart from CNN. With its promising capabilities, this approach represents a significant stride towards enhancing healthcare diagnostics and preventive measures related to vitamin deficiencies.
Synthetic Aperture Radar (SAR) has a wide range of applications in the military and civilian fields. Due to the angle sensitivity of targets in SAR images and the neglect of correlation information between images from...
详细信息
ISBN:
(数字)9798350390254
ISBN:
(纸本)9798350390261
Synthetic Aperture Radar (SAR) has a wide range of applications in the military and civilian fields. Due to the angle sensitivity of targets in SAR images and the neglect of correlation information between images from different viewpoints in single-view recognition, multi-view recognition methods have received widespread attention. We propose a multi-view rotation double-layer fusion (MVRDF) CNN-LSTM network to fuse the rotation features of multi-view images. It uses three parallel CNNs to extract features from randomly extracted multi-view images, rotates the feature sequence twice to obtain three sets of features. We add a layer of rotation angle dimension LSTM after a layer of angle dimension LSTM for double-layer fusion. The experimental results based on MSTAR and a measured dataset demonstrate the effectiveness and generalization of the proposed method.
Along with computer technology, the demand of digital imageprocessing is too high and it is used massively in every sector like organization, business, medical and so on. image segmentation enables us to analyze any ...
详细信息
Hand gesture recognition (HGR) is a hot topic in machine learning and imageprocessing communities. HGR is also vital for some Human-Computer Interaction (HCI) applications. Up to now, traditional machine learning app...
详细信息
In the robot application system incorporating dexterous hand, a vision-based robot grasping system is proposed to address the lack of robustness of dexterous hand in grasping fixed attitude objects. First, a 6DOF robo...
详细信息
ISBN:
(数字)9798350355642
ISBN:
(纸本)9798350355659
In the robot application system incorporating dexterous hand, a vision-based robot grasping system is proposed to address the lack of robustness of dexterous hand in grasping fixed attitude objects. First, a 6DOF robot grasping system based on machinevision is constructed using dexterous hand, depth camera and 6DOF collaborative robot, which realizes accurate grasping under vision guidance; second, to solve the problem of vision system's poor localization accuracy due to the loss of image information and features caused by image noise, occlusion and complex background in the process of imageprocessing, a pooling layer and attention mechanism to enhance the feature extraction ability; moreover, an optimized dexterous hand grasping strategy is proposed through exhaustive grasping action design and analysis, which effectively improves the robustness of the system. The experimental results show that the accuracy of the target detection model reaches 87% through the localization measurement of the experimental objects, which is 2.1% higher than the original method, and the grasping success rate of the robotic system equipped with dexterous hand and depth camera is improved by 3.5%. These results validate the feasibility of the robotic grasping system incorporating dexterous hands in practical applications and significantly enhance the robustness of the system.
Approximate computing has become a widely recognized method for designing energy-efficient arithmetic architectures in the context of error-tolerant applications. This paper presents the design and analysis of a 4-bit...
Approximate computing has become a widely recognized method for designing energy-efficient arithmetic architectures in the context of error-tolerant applications. This paper presents the design and analysis of a 4-bit approximate Vedic multiplier (AVMT) using the Urdhva Tiryagbhyam method. This Vedic approach, involving vertical and crosswise steps, outperforms traditional multiplication in terms of efficiency. An approximate 2-bit multiplier (AVM2) is designed, and an AVMT is proposed using AVM2. The proposed architecture has better propagation delay and less area utilization compared to other conventional multipliers. AVMT has an 11% reduction in area consumption and a 12% increase in processing speed compared to the exact Vedic multiplier. To assess its practicality in real-world scenarios, the proposed multiplier is integrated into an image-blending application. The results indicate that the system achieves a Structural Similarity Index (SSIM) average value of 0.91, which proves to be suitable for error-resilient imageprocessingapplications.
Deep learning models based on graph neural networks have emerged as a popular approach for solving computer vision problems. They encode the image into a graph structure and can be beneficial for efficiently capturing...
详细信息
Successful applications of deep learning often depend on large amount of training data. However, in practical image recognition tasks, available training data are often limited or imbalanced across classes, causing th...
详细信息
ISBN:
(纸本)9783031189098;9783031189104
Successful applications of deep learning often depend on large amount of training data. However, in practical image recognition tasks, available training data are often limited or imbalanced across classes, causing the over-fitting issue or the prediction bias issue during model training. In this paper, based on word embedding models from studies in natural language processing, the prior knowledge about the relationships between image classes is utilized to help train more generalizable classifiers under the condition of limited or class-imbalanced training data. Such inter-class relational knowledge is captured in the word embedding vectors for the textual names of image classes. Using these word embedding vectors as soft labels for corresponding image classes, the feature extractor part of a deep learning model can be guided to learn to extract visual features which contain both class-specific and class-shared information. Experiments on multiple image classification datasets confirm that the proposed learning framework helps improve model performance when training data is limited or class-imbalanced.
Among the instruments for early detection are those for analysing gases in people's faeces, as it has been found that the presence of the intensity of certain compounds is related to the presence of cancer, diabet...
详细信息
ISBN:
(数字)9783031133213
ISBN:
(纸本)9783031133213;9783031133206
Among the instruments for early detection are those for analysing gases in people's faeces, as it has been found that the presence of the intensity of certain compounds is related to the presence of cancer, diabetes or Alzheimer. The availability of sensor devices in recent years, together with the Internet of Things (IoT) paradigm, has made it possible to create low-cost systems that allow initial solutions to be tested for various real applications. Therefore, the aim of this contribution is to present the use case of a stool gas monitoring system in order to be the beginning of a solution for the early detection of this type of diseases. The proposed prototype integrates a thermal camera and MOX sensors to collect temperature and gas measurements immediately after a person has made deposition in their home. The measurements are monitored through an IoT platform and stored on a cloud server.
This article centers on foreground-background image segmentation techniques, with a particular focus on applications using two shifted or highly similar images. Our work critically examines prevalent imageprocessing ...
详细信息
ISBN:
(数字)9798350375428
ISBN:
(纸本)9798350375435
This article centers on foreground-background image segmentation techniques, with a particular focus on applications using two shifted or highly similar images. Our work critically examines prevalent imageprocessing methodologies, leading to the development and implementation of an innovative algorithm for more precise segmentation. We initially unpack the concepts of image segmentation and stereo vision, setting the stage for an in-depth exploration of various segmentation methods and techniques. Our investigation reveals that synergistic combinations of methods frequently produce more refined outcomes. We subsequently propose a practical solution, juxtaposing our approach with pre-existing alternatives to delineate its comparative strengths, weaknesses, and unique attributes.
暂无评论