首页 > 最新文献

2018 Digital Image Computing: Techniques and Applications (DICTA)最新文献

英文 中文
Ensemble Classification Technique for Water Detection in Satellite Images 卫星图像中水体检测的集成分类技术
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615870
R. Jony, A. Woodley, A. Raj, Dimitri Perrin
Satellite images are capable of providing valuable, synoptic coverage of the environment and so have been used for natural disaster assessment such as flooding. There are plenty of machine learning classifiers that can detect water in satellite images and although none are perfect they often produce acceptable results. Ensemble classifiers combine multiple classifiers and are often able to outperform their constitute classifiers. Ensemble classifiers are known to be effective for image classification in different applications but are unexplored for water detection in satellite images. This research employs an ensemble classifier to detect water in satellite images for flood assessment. Classification was performed both using individual bands and Normalized Difference Water Index (NDWI). The results show that to improve the classification accuracy with ensemble classifiers it is important to choose appropriate classifiers to ensemble. It also shows that this approach is capable of producing good classification accuracy for a seen location when bands are used and an unseen location when NDWI is used.
卫星图像能够提供有价值的环境概貌,因此已被用于评估诸如洪水之类的自然灾害。有很多机器学习分类器可以检测卫星图像中的水,尽管没有一个是完美的,但它们通常会产生可接受的结果。集成分类器结合了多个分类器,并且通常能够优于它们的构成分类器。众所周知,集成分类器在不同的应用中对图像分类是有效的,但在卫星图像中的水检测方面尚未探索。本研究采用集成分类器对卫星图像中的水进行检测,用于洪水评估。采用单个波段和归一化差水指数(NDWI)进行分类。结果表明,为了提高集成分类器的分类精度,选择合适的分类器进行集成是非常重要的。研究还表明,该方法能够在使用波段时对可见位置产生良好的分类精度,在使用NDWI时对不可见位置产生良好的分类精度。
{"title":"Ensemble Classification Technique for Water Detection in Satellite Images","authors":"R. Jony, A. Woodley, A. Raj, Dimitri Perrin","doi":"10.1109/DICTA.2018.8615870","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615870","url":null,"abstract":"Satellite images are capable of providing valuable, synoptic coverage of the environment and so have been used for natural disaster assessment such as flooding. There are plenty of machine learning classifiers that can detect water in satellite images and although none are perfect they often produce acceptable results. Ensemble classifiers combine multiple classifiers and are often able to outperform their constitute classifiers. Ensemble classifiers are known to be effective for image classification in different applications but are unexplored for water detection in satellite images. This research employs an ensemble classifier to detect water in satellite images for flood assessment. Classification was performed both using individual bands and Normalized Difference Water Index (NDWI). The results show that to improve the classification accuracy with ensemble classifiers it is important to choose appropriate classifiers to ensemble. It also shows that this approach is capable of producing good classification accuracy for a seen location when bands are used and an unseen location when NDWI is used.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125261325","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 6
Robust CNN-based Gait Verification and Identification using Skeleton Gait Energy Image 基于骨骼步态能量图像的鲁棒cnn步态验证与识别
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615802
Lingxiang Yao, Worapan Kusakunniran, Qiang Wu, Jian Zhang, Zhenmin Tang
As a kind of behavioral biometrie feature, gait has been widely applied for human verification and identification. Approaches to gait recognition can be classified into two categories: model-free approaches and model-based approaches. Model-free approaches are sensitive to appearance changes. For model-based approaches, it is difficult to extract the reliable body models from gait sequences. In this paper, based on the robust skeleton points produced from a two-branch multi-stage CNN network, a novel model-based feature, Skeleton Gait Energy Image (SGEI), has been proposed. Relevant experimental performances indicate that SGEI is more robust to the cloth changes. Another contribution is that two different CNN-based architectures have been separately proposed for gait verification and gait identification. Both these two architectures have been evaluated on the datasets. They have presented satisfying performances and increased the robustness for gait recognition in the unconstrained environments with view variances and cloth variances.
步态作为一种行为生物特征,已广泛应用于人体验证和身份识别。步态识别方法可以分为两类:无模型方法和基于模型的方法。无模型方法对外观变化很敏感。对于基于模型的方法,很难从步态序列中提取可靠的身体模型。基于两分支多阶段CNN网络生成的鲁棒骨架点,提出了一种新的基于模型的特征——骨架步态能量图像(SGEI)。相关实验结果表明,SGEI对布料变化具有较强的鲁棒性。另一个贡献是分别提出了两种不同的基于cnn的步态验证和步态识别架构。这两种架构都在数据集上进行了评估。结果表明,该方法具有较好的鲁棒性,可以在无约束的视觉和布料方差环境下进行步态识别。
{"title":"Robust CNN-based Gait Verification and Identification using Skeleton Gait Energy Image","authors":"Lingxiang Yao, Worapan Kusakunniran, Qiang Wu, Jian Zhang, Zhenmin Tang","doi":"10.1109/DICTA.2018.8615802","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615802","url":null,"abstract":"As a kind of behavioral biometrie feature, gait has been widely applied for human verification and identification. Approaches to gait recognition can be classified into two categories: model-free approaches and model-based approaches. Model-free approaches are sensitive to appearance changes. For model-based approaches, it is difficult to extract the reliable body models from gait sequences. In this paper, based on the robust skeleton points produced from a two-branch multi-stage CNN network, a novel model-based feature, Skeleton Gait Energy Image (SGEI), has been proposed. Relevant experimental performances indicate that SGEI is more robust to the cloth changes. Another contribution is that two different CNN-based architectures have been separately proposed for gait verification and gait identification. Both these two architectures have been evaluated on the datasets. They have presented satisfying performances and increased the robustness for gait recognition in the unconstrained environments with view variances and cloth variances.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127220414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 27
Face Recognition with Multi-channel Local Mesh High-order Pattern Descriptor and Convolutional Neural Network 基于多通道局部网格高阶模式描述子和卷积神经网络的人脸识别
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615831
M. Asif, Yongsheng Gao, J. Zhou
In this paper, we propose a novel Local Mesh High-order Pattern Descriptor (LMHPD) for face recognition. This description is constructed in a high-order derivative space and is integrated with a Convolutional Neural Network (CNN) architecture. Based on the information collected at a local neighborhood of reference pixel with diverse radiuses and mesh angles, a vectorized feature representation of the reference pixel is generated to provide micro-patterns. They are then converted to multi-channels to use in conjunction with the CNN. The CNN adopted in the proposed architecture is generic and very compact with a small number of convolutional layers. However, LMHPD is derived in such a way that it can work with most of the available CNN architectures. For keeping the computational cost and time complexity at the minimum, we propose a lighter approach of high-order texture descriptor with CNN architecture that can effectively extract discriminative face features. Extensive experiments on Extended Yale B and CMU-PIE datasets show that our method consistently outperforms several alternative descriptors for face recognition under various circumstances.
本文提出了一种新的局部网格高阶模式描述子(LMHPD)用于人脸识别。该描述在高阶导数空间中构建,并与卷积神经网络(CNN)架构集成。基于在参考像素的不同半径和网格角度的局部邻域收集的信息,生成参考像素的矢量化特征表示以提供微模式。然后将它们转换成多通道,与CNN一起使用。所提出的体系结构中采用的CNN是通用的,非常紧凑,卷积层数量很少。然而,LMHPD的派生方式使得它可以与大多数可用的CNN架构一起工作。为了保持最小的计算成本和时间复杂度,我们提出了一种更轻的基于CNN架构的高阶纹理描述子方法,可以有效地提取判别性人脸特征。在扩展耶鲁B和CMU-PIE数据集上的大量实验表明,我们的方法在各种情况下始终优于几种替代的人脸识别描述符。
{"title":"Face Recognition with Multi-channel Local Mesh High-order Pattern Descriptor and Convolutional Neural Network","authors":"M. Asif, Yongsheng Gao, J. Zhou","doi":"10.1109/DICTA.2018.8615831","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615831","url":null,"abstract":"In this paper, we propose a novel Local Mesh High-order Pattern Descriptor (LMHPD) for face recognition. This description is constructed in a high-order derivative space and is integrated with a Convolutional Neural Network (CNN) architecture. Based on the information collected at a local neighborhood of reference pixel with diverse radiuses and mesh angles, a vectorized feature representation of the reference pixel is generated to provide micro-patterns. They are then converted to multi-channels to use in conjunction with the CNN. The CNN adopted in the proposed architecture is generic and very compact with a small number of convolutional layers. However, LMHPD is derived in such a way that it can work with most of the available CNN architectures. For keeping the computational cost and time complexity at the minimum, we propose a lighter approach of high-order texture descriptor with CNN architecture that can effectively extract discriminative face features. Extensive experiments on Extended Yale B and CMU-PIE datasets show that our method consistently outperforms several alternative descriptors for face recognition under various circumstances.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121328189","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 2
Complexity and Entropy of Knee Kinematics in a Joint Reposition Test: Effect of Strapping and Kinesiology Taping 膝关节复位试验中膝关节运动学的复杂性和熵:绑带和运动学胶带的影响
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615777
L. Donnan, M. Paul, L. Crowley, Kilian Felesimo, H. Jelinek
Proprioception plays an important role in neuromuscular control and stability. Taping the knee or ankle with strapping tape (ST) is a common means to increase stability and limit unwanted joint motion. Kinesiology tape (KT) is an alternative tape has been proposed to enhance proprioceptive information from the skin muscles and joints. Comparisons of muscle responses associated with different taping methods has not been investigated using a joint reposition test (JRT). The current study investigated lower limb muscle activity during a blind folded JRT in a group of college students with no known injuries. Thirty nine healthy college students between 18–35 years of age were recruited using convenience sampling. Electromyographical (EMG) data was recorded from lower limb muscles and 3D video recordings tracked knee joint angle accuracy for the JRT. Participants were blindfolded and guided to 40 degrees of knee flexion by the experimenter, and were asked to repeat this joint position five times unaided. Higuchi fractal dimension and sample entropy were used to determine the nonlinear dynamic properties of the muscle responses. Statistical analysis was with repeated measures ANOVA and Tukey post hoc test. Significance was set at p<0.05. The results indicated that ST led to higher complexity and randomness of muscle activity, compared to KT. These results correlated with the absolute error associated with the JRT, where KT was significantly lower with a lower standard deviation. Higher complexity and randomness may indicate compromised muscle activity due to loss of proprioceptive information and decreased sensorimotor effectiveness.
本体感觉在神经肌肉的控制和稳定中起着重要作用。用绑带(ST)捆扎膝盖或脚踝是增加稳定性和限制不必要的关节运动的常用方法。运动机能学磁带(KT)是另一种磁带,已被提出用于增强来自皮肤肌肉和关节的本体感觉信息。使用关节复位试验(JRT)比较与不同贴带方法相关的肌肉反应尚未进行研究。目前的研究调查了一组没有已知损伤的大学生在盲人折叠JRT期间的下肢肌肉活动。采用方便抽样法,招募18 ~ 35岁的健康大学生39人。记录下肢肌肉的肌电图(EMG)数据,3D视频记录JRT的膝关节角度准确性。参与者被蒙上眼睛,在实验者的引导下屈膝40度,并被要求在没有帮助的情况下重复这个关节姿势5次。采用Higuchi分形维数和样本熵来确定肌肉响应的非线性动态特性。统计分析采用重复测量方差分析和Tukey事后检验。p<0.05为显著性。结果表明,与KT相比,ST导致肌肉活动的复杂性和随机性更高。这些结果与与JRT相关的绝对误差相关,其中KT明显较低,标准差较低。较高的复杂性和随机性可能表明由于本体感觉信息的丢失和感觉运动有效性的降低而导致肌肉活动受损。
{"title":"Complexity and Entropy of Knee Kinematics in a Joint Reposition Test: Effect of Strapping and Kinesiology Taping","authors":"L. Donnan, M. Paul, L. Crowley, Kilian Felesimo, H. Jelinek","doi":"10.1109/DICTA.2018.8615777","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615777","url":null,"abstract":"Proprioception plays an important role in neuromuscular control and stability. Taping the knee or ankle with strapping tape (ST) is a common means to increase stability and limit unwanted joint motion. Kinesiology tape (KT) is an alternative tape has been proposed to enhance proprioceptive information from the skin muscles and joints. Comparisons of muscle responses associated with different taping methods has not been investigated using a joint reposition test (JRT). The current study investigated lower limb muscle activity during a blind folded JRT in a group of college students with no known injuries. Thirty nine healthy college students between 18–35 years of age were recruited using convenience sampling. Electromyographical (EMG) data was recorded from lower limb muscles and 3D video recordings tracked knee joint angle accuracy for the JRT. Participants were blindfolded and guided to 40 degrees of knee flexion by the experimenter, and were asked to repeat this joint position five times unaided. Higuchi fractal dimension and sample entropy were used to determine the nonlinear dynamic properties of the muscle responses. Statistical analysis was with repeated measures ANOVA and Tukey post hoc test. Significance was set at p<0.05. The results indicated that ST led to higher complexity and randomness of muscle activity, compared to KT. These results correlated with the absolute error associated with the JRT, where KT was significantly lower with a lower standard deviation. Higher complexity and randomness may indicate compromised muscle activity due to loss of proprioceptive information and decreased sensorimotor effectiveness.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115566684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Dynamic Saliency Model Inspired by Middle Temporal Visual Area: A Spatio-Temporal Perspective 中时视觉区激发的动态显著性模型:一个时空视角
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615806
Hassan Mahmood, S. Islam, S. O. Gilani, Y. Ayaz
With the advancement in technology, digital visual data is also increasing day by day. And there is a great need to develop systems that can understand it. For computers, this is a daunting task to do but our brain efficiently and apparently effortlessly doing this task very well. This paper aims to devise a dynamic saliency model inspired by the human visual system. Most models are based on low-level image features and focus on static and dynamic images. And those models do not perform well in accordance with the human gaze movement for dynamic scenes. We here demonstrate that a combined model of bio-inspired spatio-temporal features, high-level and low-level features outperform listed models in predicting human fixation on dynamic visual input. Our comparison with other models is based on eye-movement recordings of human participants observing dynamic natural scenes.
随着技术的进步,数字视觉数据也日益增多。我们非常需要开发能够理解它的系统。对于计算机来说,这是一项艰巨的任务,但我们的大脑却能高效且毫不费力地完成这项任务。本文旨在设计一个受人类视觉系统启发的动态显著性模型。大多数模型基于底层图像特征,关注静态和动态图像。在动态场景下,这些模型不能很好地反映人类的注视运动。我们在此证明了生物启发时空特征、高级和低级特征的组合模型在预测人类对动态视觉输入的注视方面优于列表模型。我们与其他模型的比较是基于观察动态自然场景的人类参与者的眼动记录。
{"title":"Dynamic Saliency Model Inspired by Middle Temporal Visual Area: A Spatio-Temporal Perspective","authors":"Hassan Mahmood, S. Islam, S. O. Gilani, Y. Ayaz","doi":"10.1109/DICTA.2018.8615806","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615806","url":null,"abstract":"With the advancement in technology, digital visual data is also increasing day by day. And there is a great need to develop systems that can understand it. For computers, this is a daunting task to do but our brain efficiently and apparently effortlessly doing this task very well. This paper aims to devise a dynamic saliency model inspired by the human visual system. Most models are based on low-level image features and focus on static and dynamic images. And those models do not perform well in accordance with the human gaze movement for dynamic scenes. We here demonstrate that a combined model of bio-inspired spatio-temporal features, high-level and low-level features outperform listed models in predicting human fixation on dynamic visual input. Our comparison with other models is based on eye-movement recordings of human participants observing dynamic natural scenes.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131104584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Deep Learning on Brain Cortical Thickness Data for Disease Classification 脑皮质厚度数据的深度学习用于疾病分类
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615775
Medhani Menikdiwela, Chuong V. Nguyen, Marnie E. Shaw
Deep learning has been applied to learn and classify brain disease using volumetric MRI scans with an accuracy approaching or even exceeding that of a human expert. This is typically done by applying convolutional neural networks to slices of a 3D brain image volume. Each slice of the brain volume, however, represents only a small cross-sectional area of the cortical layer. On the other hand, convolutional neural networks are less well developed for 3D volumes. Therefore we sought to apply deep networks to the 2D cortical surface, for the purpose of classifying Alzheimer's disease (AD). AD is known to affect the thickness and geometry of the cortical surface of the brain. Although the cortical surface has a complex geometry, here we present a novel data processing method to feed the information of an entire cortical surface into existing deep networks for more accurate early disease detection. A brain 3D MRI volume is registered and its cortical surface is flattened to a 2D plane. The flattened distributions of the thickness, curvature and surface area are combined into an RBG image which can be readily fed to existing deep networks. In this paper, the ADNI dataset of brain MRI scans are used and flattened cortical images are applied to different deep networks including ResNet and Inception. Two pre-clinical stages of AD are considered; stable mild cognitive impairment (MCIs) and converting mild cognitive impairment (MCIc). Experiments show that using flattened cortical images consistently leads to higher accuracy compared to using brain slices with the same network architecture. Specifically, the highest accuracy of 81% is achieved by Inception with flattened cortical images, as compared to 68% by the same network on brain slices and 75.9% accuracy by the best method in the literature which also used a deep network on brain slices. Our results indicate that flattened cortical images can be used to learn and classify AD with high accuracy.
深度学习已被应用于通过体积核磁共振扫描来学习和分类脑部疾病,其准确性接近甚至超过人类专家。这通常是通过将卷积神经网络应用于3D脑图像体积的切片来完成的。然而,每块脑体积切片只代表皮质层的一小块横截面积。另一方面,卷积神经网络在3D体积上的发展并不好。因此,我们试图将深度网络应用于二维皮层表面,以对阿尔茨海默病(AD)进行分类。已知阿尔茨海默病会影响大脑皮层表面的厚度和几何形状。尽管皮质表面具有复杂的几何结构,但我们提出了一种新的数据处理方法,将整个皮质表面的信息馈送到现有的深度网络中,以更准确地进行早期疾病检测。脑三维MRI体积登记,其皮质表面被平展到二维平面。厚度、曲率和表面积的扁平分布被组合成一个RBG图像,可以很容易地馈送到现有的深度网络。本文使用脑MRI扫描的ADNI数据集,并将平面化的皮质图像应用于不同的深度网络,包括ResNet和Inception。阿尔茨海默病的两个临床前阶段被考虑;稳定型轻度认知障碍(MCIs)和转换型轻度认知障碍(MCIc)。实验表明,与使用具有相同网络结构的大脑切片相比,始终使用平坦的皮层图像具有更高的准确性。具体来说,盗梦空间使用平坦的皮质图像获得了81%的最高准确率,相比之下,相同的网络在大脑切片上获得了68%的准确率,而文献中使用深度网络在大脑切片上获得的最佳方法的准确率为75.9%。我们的研究结果表明,平坦的皮质图像可以用于AD的学习和分类,准确率很高。
{"title":"Deep Learning on Brain Cortical Thickness Data for Disease Classification","authors":"Medhani Menikdiwela, Chuong V. Nguyen, Marnie E. Shaw","doi":"10.1109/DICTA.2018.8615775","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615775","url":null,"abstract":"Deep learning has been applied to learn and classify brain disease using volumetric MRI scans with an accuracy approaching or even exceeding that of a human expert. This is typically done by applying convolutional neural networks to slices of a 3D brain image volume. Each slice of the brain volume, however, represents only a small cross-sectional area of the cortical layer. On the other hand, convolutional neural networks are less well developed for 3D volumes. Therefore we sought to apply deep networks to the 2D cortical surface, for the purpose of classifying Alzheimer's disease (AD). AD is known to affect the thickness and geometry of the cortical surface of the brain. Although the cortical surface has a complex geometry, here we present a novel data processing method to feed the information of an entire cortical surface into existing deep networks for more accurate early disease detection. A brain 3D MRI volume is registered and its cortical surface is flattened to a 2D plane. The flattened distributions of the thickness, curvature and surface area are combined into an RBG image which can be readily fed to existing deep networks. In this paper, the ADNI dataset of brain MRI scans are used and flattened cortical images are applied to different deep networks including ResNet and Inception. Two pre-clinical stages of AD are considered; stable mild cognitive impairment (MCIs) and converting mild cognitive impairment (MCIc). Experiments show that using flattened cortical images consistently leads to higher accuracy compared to using brain slices with the same network architecture. Specifically, the highest accuracy of 81% is achieved by Inception with flattened cortical images, as compared to 68% by the same network on brain slices and 75.9% accuracy by the best method in the literature which also used a deep network on brain slices. Our results indicate that flattened cortical images can be used to learn and classify AD with high accuracy.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128503283","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 16
Massively Parallel Implementation of a Fast Resource Efficient White Light Interferometry Algorithm 一种快速资源高效白光干涉测量算法的大规模并行实现
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615828
Tobias Scholz, M. Rosenberger, G. Notni
In this paper an implementation of a massively parallel white light interferometry algorithm will be presented. In contrast to more common algorithms it not depends on the fast Fourier transform. Using non-equidistant sampling steps is supported and will occur after compression. The algorithm can be applied to variety of target hardware ranging from embedded implementations with limited resources up to desktop computers and higher. It was invented to use the massively parallel architecture of field-programmable gate arrays (FPGA). The approach was proven on the Xilinx Zynq architecture and an x86 high level language implementation. Major improvements compared to more common solutions was the ability to compress the raw data easily while keeping the accuracy despite the limited hardware resources available. Independent of the height of the raw image stack the reconstruction can be solved in constant time.
本文将提出一种大规模平行白光干涉测量算法的实现。与更常见的算法相比它不依赖于快速傅里叶变换。支持使用非等距采样步骤,并将在压缩后发生。该算法可以应用于各种目标硬件,从资源有限的嵌入式实现到台式计算机和更高的硬件。它的发明是为了利用现场可编程门阵列(FPGA)的大规模并行架构。该方法在Xilinx Zynq架构和x86高级语言实现上得到了验证。与更常见的解决方案相比,主要的改进是能够轻松压缩原始数据,同时在硬件资源有限的情况下保持准确性。该方法不受原始图像叠加高度的影响,可以在恒定时间内完成重构。
{"title":"Massively Parallel Implementation of a Fast Resource Efficient White Light Interferometry Algorithm","authors":"Tobias Scholz, M. Rosenberger, G. Notni","doi":"10.1109/DICTA.2018.8615828","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615828","url":null,"abstract":"In this paper an implementation of a massively parallel white light interferometry algorithm will be presented. In contrast to more common algorithms it not depends on the fast Fourier transform. Using non-equidistant sampling steps is supported and will occur after compression. The algorithm can be applied to variety of target hardware ranging from embedded implementations with limited resources up to desktop computers and higher. It was invented to use the massively parallel architecture of field-programmable gate arrays (FPGA). The approach was proven on the Xilinx Zynq architecture and an x86 high level language implementation. Major improvements compared to more common solutions was the ability to compress the raw data easily while keeping the accuracy despite the limited hardware resources available. Independent of the height of the raw image stack the reconstruction can be solved in constant time.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"76 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121714785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 1
Classification of White Blood Cells using Bispectral Invariant Features of Nuclei Shape 利用细胞核形状的双谱不变性特征对白细胞进行分类
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615762
Khamael Al-Dulaimi, V. Chandran, Jasmine Banks, Inmaculada Tomeo-Reyes, Kien Nguyen
Classification of white blood cells from microscope images is a challenging task, especially in the choice of feature representation, considering intra-class variations arising from non-uniform illumination, stage of maturity, scale, rotation and shifting. In this paper, we propose a new feature extraction scheme relying on bispectral invariant features which are robust to these challenges. Bispectral invariant features are extracted from the shape of segmented white blood cell nuclei. Segmentation of white blood cell nuclei is achieved using a level set algorithm via geometric active contours. Binary support vector machines and a classification tree are used for classifying multiple classes of the cells. Performance of the proposed method is evaluated on a combined dataset of 10 classes with 460 white blood cell images collected from 3 datasets and using 5-fold cross validation. It achieves an average classification accuracy of 96.13% and outperforms other popular representations including local binary pattern, histogram of oriented gradients, local directional pattern and speeded up robust features with the same classifier over the same data. The classification accuracy of the proposed method is also compared and benchmarked with the other existing techniques for classification white blood cells into 10 classes over the same datasets and the results show that the proposed method is superior over other approaches.
从显微镜图像中对白细胞进行分类是一项具有挑战性的任务,特别是在特征表示的选择上,考虑到不均匀光照、成熟阶段、尺度、旋转和移动等引起的类内变化。在本文中,我们提出了一种新的基于双谱不变特征的特征提取方案,该方案对这些挑战具有鲁棒性。从分割的白细胞细胞核形状中提取双谱不变性特征。利用水平集算法通过几何活动轮廓实现了白细胞细胞核的分割。使用二值支持向量机和分类树对多类细胞进行分类。采用5倍交叉验证的方法,对从3个数据集收集的460张白细胞图像的10类组合数据集进行了性能评估。它的平均分类准确率达到96.13%,优于其他常用的表示方法,包括局部二值模式、定向梯度直方图、局部方向模式,并在相同的分类器上加速了相同数据的鲁棒特征。在相同的数据集上,将该方法的分类精度与其他现有的将白细胞分为10类的方法进行了比较和基准测试,结果表明该方法优于其他方法。
{"title":"Classification of White Blood Cells using Bispectral Invariant Features of Nuclei Shape","authors":"Khamael Al-Dulaimi, V. Chandran, Jasmine Banks, Inmaculada Tomeo-Reyes, Kien Nguyen","doi":"10.1109/DICTA.2018.8615762","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615762","url":null,"abstract":"Classification of white blood cells from microscope images is a challenging task, especially in the choice of feature representation, considering intra-class variations arising from non-uniform illumination, stage of maturity, scale, rotation and shifting. In this paper, we propose a new feature extraction scheme relying on bispectral invariant features which are robust to these challenges. Bispectral invariant features are extracted from the shape of segmented white blood cell nuclei. Segmentation of white blood cell nuclei is achieved using a level set algorithm via geometric active contours. Binary support vector machines and a classification tree are used for classifying multiple classes of the cells. Performance of the proposed method is evaluated on a combined dataset of 10 classes with 460 white blood cell images collected from 3 datasets and using 5-fold cross validation. It achieves an average classification accuracy of 96.13% and outperforms other popular representations including local binary pattern, histogram of oriented gradients, local directional pattern and speeded up robust features with the same classifier over the same data. The classification accuracy of the proposed method is also compared and benchmarked with the other existing techniques for classification white blood cells into 10 classes over the same datasets and the results show that the proposed method is superior over other approaches.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114387637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 26
A Scale-Free and Parameter-Free Image Edge Strength Measure 一种无尺度、无参数的图像边缘强度测量方法
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615813
Guy Smith, P. Jackway
We present a family of image Slope Measures which are scale-free measures that are highest near image edges. They are defined at each pixel as the steepest of the (up, down, or bi-directional) intensity slopes to every other pixel. We list some useful mathematical properties such as intensity and rotation invariances and show a relationship to the maximal morphological dilations and erosions by cones. We discuss generalisations by using non-Euclidean distances or non-conical structuring functions, and extensions to colour, multi-spectral and higher-dimensional images. We present detailed pseudo-code for a fast doubly-recursive multi-resolution algorithm and give typical algorithm timings and visually demonstrate the measure as applied to standard test images. Reference C code for these algorithms is available on the internet at: https://github.com/xomexx/SlopeMeasures.
我们提出了一组图像斜率测度,它们是图像边缘附近最高的无尺度测度。它们在每个像素上被定义为每一个其他像素的最陡(向上、向下或双向)强度斜率。我们列出了一些有用的数学性质,如强度和旋转不变性,并显示了锥的最大形态扩张和侵蚀的关系。我们通过使用非欧几里得距离或非圆锥结构函数来讨论泛化,并扩展到彩色,多光谱和高维图像。我们给出了一种快速双递归多分辨率算法的详细伪代码,给出了典型的算法时序,并直观地演示了该方法在标准测试图像上的应用。这些算法的参考C代码可以在互联网上找到:https://github.com/xomexx/SlopeMeasures。
{"title":"A Scale-Free and Parameter-Free Image Edge Strength Measure","authors":"Guy Smith, P. Jackway","doi":"10.1109/DICTA.2018.8615813","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615813","url":null,"abstract":"We present a family of image Slope Measures which are scale-free measures that are highest near image edges. They are defined at each pixel as the steepest of the (up, down, or bi-directional) intensity slopes to every other pixel. We list some useful mathematical properties such as intensity and rotation invariances and show a relationship to the maximal morphological dilations and erosions by cones. We discuss generalisations by using non-Euclidean distances or non-conical structuring functions, and extensions to colour, multi-spectral and higher-dimensional images. We present detailed pseudo-code for a fast doubly-recursive multi-resolution algorithm and give typical algorithm timings and visually demonstrate the measure as applied to standard test images. Reference C code for these algorithms is available on the internet at: https://github.com/xomexx/SlopeMeasures.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124448700","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Active Contours Based on An Anisotropic Diffusion 基于各向异性扩散的活动轮廓
Pub Date : 2018-12-01 DOI: 10.1109/DICTA.2018.8615767
Shafiullah Soomro, K. Choi
Image Segmentation is one of the pivotal procedure in the field of imaging and its objective is to catch required boundaries inside an image. In this paper, we propose a novel active contour method based on anisotropic diffusion. Global regionbased active contour methods rely on global intensity information across the regions. However, these methods fail to produce desired segmentation results when an image has some background variations or noise. In this regard, we adapt Perona and Malik smoothing technique as enhancement step. This technique provides interregional smoothing, sharpens the boundaries and blurs the background of an image. Our main role is the formulation of a new SPF (signed pressure force) function, which uses global intensity information across the regions. Minimizing an energy function using partial differential framework produce results with semantically meaningful boundaries instead of capturing impassive regions. Finally, we use Gaussian kernel to eliminate problem of reinitialization in level set function. We use images taken from different modalities to validate the outcome of the proposed method. In the result section, we have evaluated that, the proposed method achieves good results qualitatively and quantitatively with high accuracy compared to other state-of-the-art models.
图像分割是成像领域的关键步骤之一,其目的是捕获图像内部所需的边界。本文提出了一种基于各向异性扩散的活动轮廓线方法。基于区域的全球活动等高线方法依赖于区域间的全球强度信息。然而,当图像有背景变化或噪声时,这些方法无法产生理想的分割结果。在这方面,我们采用Perona和Malik平滑技术作为增强步骤。这种技术提供了区域间平滑,锐化边界和模糊图像的背景。我们的主要作用是制定一个新的SPF(签名压力)函数,它使用了各个地区的全球强度信息。使用偏微分框架最小化能量函数产生具有语义上有意义的边界的结果,而不是捕获无表情区域。最后,利用高斯核消除了水平集函数的重新初始化问题。我们使用不同模式的图像来验证所提出方法的结果。在结果部分,我们已经评估了,与其他最先进的模型相比,所提出的方法在定性和定量方面取得了良好的结果,精度很高。
{"title":"Active Contours Based on An Anisotropic Diffusion","authors":"Shafiullah Soomro, K. Choi","doi":"10.1109/DICTA.2018.8615767","DOIUrl":"https://doi.org/10.1109/DICTA.2018.8615767","url":null,"abstract":"Image Segmentation is one of the pivotal procedure in the field of imaging and its objective is to catch required boundaries inside an image. In this paper, we propose a novel active contour method based on anisotropic diffusion. Global regionbased active contour methods rely on global intensity information across the regions. However, these methods fail to produce desired segmentation results when an image has some background variations or noise. In this regard, we adapt Perona and Malik smoothing technique as enhancement step. This technique provides interregional smoothing, sharpens the boundaries and blurs the background of an image. Our main role is the formulation of a new SPF (signed pressure force) function, which uses global intensity information across the regions. Minimizing an energy function using partial differential framework produce results with semantically meaningful boundaries instead of capturing impassive regions. Finally, we use Gaussian kernel to eliminate problem of reinitialization in level set function. We use images taken from different modalities to validate the outcome of the proposed method. In the result section, we have evaluated that, the proposed method achieves good results qualitatively and quantitatively with high accuracy compared to other state-of-the-art models.","PeriodicalId":130057,"journal":{"name":"2018 Digital Image Computing: Techniques and Applications (DICTA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132544227","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 4
期刊
2018 Digital Image Computing: Techniques and Applications (DICTA)
全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1