Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470367
Yuhao Chen, Javier Ribera, E. Delp
Phenotyping is the process of estimating the physical and chemical properties of a plant. Traditional phenotyping is labor intensive and time consuming. These measurements can be obtained faster by collecting aerial images with an Unmanned Aerial Vehicle (UAV) and analyzing them using modern image analysis technologies. We propose a method to estimate plant centers by classifying each pixel as a plant center or not a plant center. We then label the center of each cluster as the plant location. We studied the performance of our method on two datasets. We achieved 84% precision and 90% recall on one dataset consisting of early stage plants and 62% precision and 77% recall on another dataset consisting of later stage plants.
{"title":"Estimating Plant Centers Using A Deep Binary Classifier","authors":"Yuhao Chen, Javier Ribera, E. Delp","doi":"10.1109/SSIAI.2018.8470367","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470367","url":null,"abstract":"Phenotyping is the process of estimating the physical and chemical properties of a plant. Traditional phenotyping is labor intensive and time consuming. These measurements can be obtained faster by collecting aerial images with an Unmanned Aerial Vehicle (UAV) and analyzing them using modern image analysis technologies. We propose a method to estimate plant centers by classifying each pixel as a plant center or not a plant center. We then label the center of each cluster as the plant location. We studied the performance of our method on two datasets. We achieved 84% precision and 90% recall on one dataset consisting of early stage plants and 62% precision and 77% recall on another dataset consisting of later stage plants.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127387521","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470323
J. Choe, D. M. Montserrat, A. Schwichtenberg, E. Delp
Videosomnography (VSG) is a range of video-based methods used to record and assess sleep vs. wake states in adults and children. Traditional behavioral-VSG (B-VSG) coding requires almost real-time visual inspection by a trained technicians/coders to determine sleep vs wake states. In this paper we describe an automated VSG sleep detection system (auto-VSG) which employs motion analysis to determine sleep vs. wake states in young children. We used child head size to normalize the motion index and to provide an individual motion maximum for each child. We compared the proposed auto-VSG method to (1) traditional B-VSG codes and (2) actigraphy sleep vs. wake estimates across four sleep parameters: sleep onset time, sleep offset time, awake duration, and sleep duration. In sum, analyses revealed that estimates generated from the proposed auto-VSG method and B-VSG are comparable.
{"title":"Sleep Analysis Using Motion and Head Detection","authors":"J. Choe, D. M. Montserrat, A. Schwichtenberg, E. Delp","doi":"10.1109/SSIAI.2018.8470323","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470323","url":null,"abstract":"Videosomnography (VSG) is a range of video-based methods used to record and assess sleep vs. wake states in adults and children. Traditional behavioral-VSG (B-VSG) coding requires almost real-time visual inspection by a trained technicians/coders to determine sleep vs wake states. In this paper we describe an automated VSG sleep detection system (auto-VSG) which employs motion analysis to determine sleep vs. wake states in young children. We used child head size to normalize the motion index and to provide an individual motion maximum for each child. We compared the proposed auto-VSG method to (1) traditional B-VSG codes and (2) actigraphy sleep vs. wake estimates across four sleep parameters: sleep onset time, sleep offset time, awake duration, and sleep duration. In sum, analyses revealed that estimates generated from the proposed auto-VSG method and B-VSG are comparable.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121619656","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470356
Nolang Fanani, R. Mester
We analyze the depth reconstruction precision and sensitivity of two-frame triangulation for the case of general motion, and focus on the case of monocular visual odometry, that is: a single camera looking mostly in the direction of motion. The results confirm intuitive assumptions about the limited triangulation precision close to the focus of expansion.
{"title":"The precision of triangulation in monocular visual odometry","authors":"Nolang Fanani, R. Mester","doi":"10.1109/SSIAI.2018.8470356","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470356","url":null,"abstract":"We analyze the depth reconstruction precision and sensitivity of two-frame triangulation for the case of general motion, and focus on the case of monocular visual odometry, that is: a single camera looking mostly in the direction of motion. The results confirm intuitive assumptions about the limited triangulation precision close to the focus of expansion.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132414310","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470308
Zeina Sinno, A. Bovik
The visual brain is optimally designed to process images from the natural environment that we perceive. Describing the natural environment statistically helps in understanding how the brain encodes those images efficiently. The Natural Scene Statistics (NSS) of the luminance component of images is the basis of several univariate and bivariate statistical models. The NSS of other colors or chromatic components have been less well-analyzed. In this paper, we study the univariate and bivariate NSS of luminance and other chromatic components and how they relate.
{"title":"On the Natural Statistics of Chromatic Images","authors":"Zeina Sinno, A. Bovik","doi":"10.1109/SSIAI.2018.8470308","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470308","url":null,"abstract":"The visual brain is optimally designed to process images from the natural environment that we perceive. Describing the natural environment statistically helps in understanding how the brain encodes those images efficiently. The Natural Scene Statistics (NSS) of the luminance component of images is the basis of several univariate and bivariate statistical models. The NSS of other colors or chromatic components have been less well-analyzed. In this paper, we study the univariate and bivariate NSS of luminance and other chromatic components and how they relate.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"246 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134499808","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470331
A. Jacoby, M. Pattichis, Sylvia Celedón-Pattichis, Carlos A. LópezLeiva
Human activity classification remains challenging due to the strong need to eliminate structural noise, the multitude of possible activities, and the strong variations in video acquisition. The current paper explores the study of human activity classification in a collaborative learning environment.This paper explores the use of color based object detection in conjunction with contextualization of object interaction to isolate motion vectors specific to each human activity. The basic approach is to make use of separate classifiers for each activity. Here, we consider the detection of typing, writing, and talking activities in raw videos.The method was tested using 43 uncropped video clips with 620 video frames for writing, 1050 for typing, and 1755 frames for talking. Using simple KNN classifiers, the method gave accuracies of 72.6% for writing, 71% for typing and 84.6% for talking. Classification accuracy improved to 92.5% (writing), 82.5% (typing) and 99.7% (talking) with the use of Deep Neural Networks.
{"title":"Context-Sensitive Human Activity Classification in Collaborative Learning Environments","authors":"A. Jacoby, M. Pattichis, Sylvia Celedón-Pattichis, Carlos A. LópezLeiva","doi":"10.1109/SSIAI.2018.8470331","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470331","url":null,"abstract":"Human activity classification remains challenging due to the strong need to eliminate structural noise, the multitude of possible activities, and the strong variations in video acquisition. The current paper explores the study of human activity classification in a collaborative learning environment.This paper explores the use of color based object detection in conjunction with contextualization of object interaction to isolate motion vectors specific to each human activity. The basic approach is to make use of separate classifiers for each activity. Here, we consider the detection of typing, writing, and talking activities in raw videos.The method was tested using 43 uncropped video clips with 620 video frames for writing, 1050 for typing, and 1755 frames for talking. Using simple KNN classifiers, the method gave accuracies of 72.6% for writing, 71% for typing and 84.6% for talking. Classification accuracy improved to 92.5% (writing), 82.5% (typing) and 99.7% (talking) with the use of Deep Neural Networks.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132777854","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470321
Xiangyu Liu, Hua Xie, B. Nutter, S. Mitra
Over the years, resting state functional magnetic resonance imaging (rsfMRI) has been a preferred design tool to analyze human brain functions and brain parcellations. Several different statistical methods have been proposed to study functional connectivity and generate various parcellation atlases based on corresponding connectivity maps. In this study, we employ a sliding window correlation method to generate accurate individual voxel-wise dynamic functional connectivity maps, based on which the brain can be parcellated into highly homogeneous functional parcels. Because there is no ground truth for functional brain parcellation, we accomplish parcellation via k-means clustering to compare with other available parcellations. With temporal characteristics of functional connectivity taken into consideration, high homogeneity can be observed in high resolution parcellation of human brain.
{"title":"High-homogeneity functional parcellation of human brain for investigating robust functional connectivity","authors":"Xiangyu Liu, Hua Xie, B. Nutter, S. Mitra","doi":"10.1109/SSIAI.2018.8470321","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470321","url":null,"abstract":"Over the years, resting state functional magnetic resonance imaging (rsfMRI) has been a preferred design tool to analyze human brain functions and brain parcellations. Several different statistical methods have been proposed to study functional connectivity and generate various parcellation atlases based on corresponding connectivity maps. In this study, we employ a sliding window correlation method to generate accurate individual voxel-wise dynamic functional connectivity maps, based on which the brain can be parcellated into highly homogeneous functional parcels. Because there is no ground truth for functional brain parcellation, we accomplish parcellation via k-means clustering to compare with other available parcellations. With temporal characteristics of functional connectivity taken into consideration, high homogeneity can be observed in high resolution parcellation of human brain.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133010714","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470377
Rohit C. Philip, Sree Ramya S. P. Malladi, M. Niihori, A. Jacob, Jeffrey J. Rodríguez
Supervised machine learning schemes are widely used to perform classification tasks. There is a wide variety of classifiers in use today, such as single- and multi-class support vector machines, k-nearest neighbors, decision trees, random forests, naive Bayes classifiers with or without kernel density estimation, linear discriminant analysis, quadratic discriminant analysis, and numerous neural network architectures. Our prior work used high-level shape, intensity, and texture features as predictors in a single-class support vector machine classifier to classify images of zebrafish neuromasts obtained using confocal microscopy into four discrete damage classes. Here, we analyze the performance of a multitude of supervised classifiers in terms of mean absolute error using these high-level features as predictors. In addition, we also analyze performance while using raw pixel data as predictors.
{"title":"Performance of Supervised Classifiers for Damage Scoring of Zebrafish Neuromasts","authors":"Rohit C. Philip, Sree Ramya S. P. Malladi, M. Niihori, A. Jacob, Jeffrey J. Rodríguez","doi":"10.1109/SSIAI.2018.8470377","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470377","url":null,"abstract":"Supervised machine learning schemes are widely used to perform classification tasks. There is a wide variety of classifiers in use today, such as single- and multi-class support vector machines, k-nearest neighbors, decision trees, random forests, naive Bayes classifiers with or without kernel density estimation, linear discriminant analysis, quadratic discriminant analysis, and numerous neural network architectures. Our prior work used high-level shape, intensity, and texture features as predictors in a single-class support vector machine classifier to classify images of zebrafish neuromasts obtained using confocal microscopy into four discrete damage classes. Here, we analyze the performance of a multitude of supervised classifiers in terms of mean absolute error using these high-level features as predictors. In addition, we also analyze performance while using raw pixel data as predictors.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131608591","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}
Pub Date : 2018-04-08DOI: 10.1109/SSIAI.2018.8470318
Gabriel Salvador, Juan M. Chau, Jorge Quesada, Cesar Carranza
Median filtering has become a ubiquitous smoothing tool for image denoising tasks, with its complexity generally determined by the median algorithm used (usually on the order of O(n log(n)) when computing the median of n elements). Most algorithms were formulated for scalar single processor computers, with few of them successfully adapted and implemented for computers with a parallel architecture. However, the redundancy for processing neighboring pixels has not yet been fully exploited for parallel implementations. Additionally, most of the implementations are only suitable for fixed point images, but not for floating point.In this paper we propose an efficient parallel implementation of the 2D median filter, based on a multiple pixel-per-thread framework, and test its implementation on a CUDA-capable GPU either for fixed point or floating point data. Our computational results show that our proposed methods outperforms state-of the art implementations, with the difference increasing significantly as the filter size grows.
{"title":"Efficient GPU-based implementation of the median filter based on a multi-pixel-per-thread framework","authors":"Gabriel Salvador, Juan M. Chau, Jorge Quesada, Cesar Carranza","doi":"10.1109/SSIAI.2018.8470318","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470318","url":null,"abstract":"Median filtering has become a ubiquitous smoothing tool for image denoising tasks, with its complexity generally determined by the median algorithm used (usually on the order of O(n log(n)) when computing the median of n elements). Most algorithms were formulated for scalar single processor computers, with few of them successfully adapted and implemented for computers with a parallel architecture. However, the redundancy for processing neighboring pixels has not yet been fully exploited for parallel implementations. Additionally, most of the implementations are only suitable for fixed point images, but not for floating point.In this paper we propose an efficient parallel implementation of the 2D median filter, based on a multiple pixel-per-thread framework, and test its implementation on a CUDA-capable GPU either for fixed point or floating point data. Our computational results show that our proposed methods outperforms state-of the art implementations, with the difference increasing significantly as the filter size grows.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125870840","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}
Pub Date : 2018-04-01DOI: 10.1109/SSIAI.2018.8470374
Mohammed Abdul Rahman, T. Zayed
Ground Penetrating Radar (GPR) is widely utilized as a Non-destructive technique by transportation authorities for inspection of bridge decks due to its ability to identify major subsurface defects in a short span of time. The attenuation of recorded signal at rebar level form a characteristic hyperbolic shape in profiles obtained from GPR scans and corresponds to the corrosiveness state of concrete. The detection of these hyperbolic regions is of paramount importance and is a precursor to successful interpretation of GPR data. This paper aims to automate the detection of hyperbolic regions or hyperbolas in GPR profiles based on Viola-Jones Algorithm. A custom detector is obtained through training with numerous samples of hyperbolas over multiple stages. The detection is achieved through the developed detector and it was applied over a complete bridge deck for validation purpose. The eventual goal of such detection is to facilitate the automation of GPR data analysis.
{"title":"Viola-Jones Algorithm for Automatic Detection of Hyperbolic Regions in GPR Profiles of Bridge Decks","authors":"Mohammed Abdul Rahman, T. Zayed","doi":"10.1109/SSIAI.2018.8470374","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470374","url":null,"abstract":"Ground Penetrating Radar (GPR) is widely utilized as a Non-destructive technique by transportation authorities for inspection of bridge decks due to its ability to identify major subsurface defects in a short span of time. The attenuation of recorded signal at rebar level form a characteristic hyperbolic shape in profiles obtained from GPR scans and corresponds to the corrosiveness state of concrete. The detection of these hyperbolic regions is of paramount importance and is a precursor to successful interpretation of GPR data. This paper aims to automate the detection of hyperbolic regions or hyperbolas in GPR profiles based on Viola-Jones Algorithm. A custom detector is obtained through training with numerous samples of hyperbolas over multiple stages. The detection is achieved through the developed detector and it was applied over a complete bridge deck for validation purpose. The eventual goal of such detection is to facilitate the automation of GPR data analysis.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116829782","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}
Pub Date : 2018-04-01DOI: 10.1109/SSIAI.2018.8470343
S. Yarlagadda, F. Zhu
Shadows are common aspect of images and when left undetected can hinder scene understanding and visual processing. We propose a simple yet effective approach based on reflectance to detect shadows from single image. An image is first segmented and based on the reflectance, illumination and texture characteristics, segments pairs are identified as shadow and non-shadow pairs. The proposed method is tested on two publicly available and widely used datasets. Our method achieves higher accuracy in detecting shadows compared to previous reported methods despite requiring fewer parameters. We also show results of shadow-free images by relighting the pixels in the detected shadow regions.
{"title":"A Reflectance Based Method For Shadow Detection and Removal","authors":"S. Yarlagadda, F. Zhu","doi":"10.1109/SSIAI.2018.8470343","DOIUrl":"https://doi.org/10.1109/SSIAI.2018.8470343","url":null,"abstract":"Shadows are common aspect of images and when left undetected can hinder scene understanding and visual processing. We propose a simple yet effective approach based on reflectance to detect shadows from single image. An image is first segmented and based on the reflectance, illumination and texture characteristics, segments pairs are identified as shadow and non-shadow pairs. The proposed method is tested on two publicly available and widely used datasets. Our method achieves higher accuracy in detecting shadows compared to previous reported methods despite requiring fewer parameters. We also show results of shadow-free images by relighting the pixels in the detected shadow regions.","PeriodicalId":422209,"journal":{"name":"2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132977132","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：