Matching Items (6)
Filtering by
- All Subjects: Computer vision
- Creators: Frakes, David
- Resource Type: Text
- Status: Published
Description
Image understanding has been playing an increasingly crucial role in vision applications. Sparse models form an important component in image understanding, since the statistics of natural images reveal the presence of sparse structure. Sparse methods lead to parsimonious models, in addition to being efficient for large scale learning. In sparse modeling, data is represented as a sparse linear combination of atoms from a "dictionary" matrix. This dissertation focuses on understanding different aspects of sparse learning, thereby enhancing the use of sparse methods by incorporating tools from machine learning. With the growing need to adapt models for large scale data, it is important to design dictionaries that can model the entire data space and not just the samples considered. By exploiting the relation of dictionary learning to 1-D subspace clustering, a multilevel dictionary learning algorithm is developed, and it is shown to outperform conventional sparse models in compressed recovery, and image denoising. Theoretical aspects of learning such as algorithmic stability and generalization are considered, and ensemble learning is incorporated for effective large scale learning. In addition to building strategies for efficiently implementing 1-D subspace clustering, a discriminative clustering approach is designed to estimate the unknown mixing process in blind source separation. By exploiting the non-linear relation between the image descriptors, and allowing the use of multiple features, sparse methods can be made more effective in recognition problems. The idea of multiple kernel sparse representations is developed, and algorithms for learning dictionaries in the feature space are presented. Using object recognition experiments on standard datasets it is shown that the proposed approaches outperform other sparse coding-based recognition frameworks. Furthermore, a segmentation technique based on multiple kernel sparse representations is developed, and successfully applied for automated brain tumor identification. Using sparse codes to define the relation between data samples can lead to a more robust graph embedding for unsupervised clustering. By performing discriminative embedding using sparse coding-based graphs, an algorithm for measuring the glomerular number in kidney MRI images is developed. Finally, approaches to build dictionaries for local sparse coding of image descriptors are presented, and applied to object recognition and image retrieval.
ContributorsJayaraman Thiagarajan, Jayaraman (Author) / Spanias, Andreas (Thesis advisor) / Frakes, David (Committee member) / Tepedelenlioğlu, Cihan (Committee member) / Turaga, Pavan (Committee member) / Arizona State University (Publisher)
Created2013
Description
There is a growing interest for improved high-accuracy camera calibration methods due to the increasing demand for 3D visual media in commercial markets. Camera calibration is used widely in the fields of computer vision, robotics and 3D reconstruction. Camera calibration is the first step for extracting 3D data from a 2D image. It plays a crucial role in computer vision and 3D reconstruction due to the fact that the accuracy of the reconstruction and 3D coordinate determination relies on the accuracy of the camera calibration to a great extent. This thesis presents a novel camera calibration method using a circular calibration pattern. The disadvantages and issues with existing state-of-the-art methods are discussed and are overcome in this work. The implemented system consists of techniques of local adaptive segmentation, ellipse fitting, projection and optimization. Simulation results are presented to illustrate the performance of the proposed scheme. These results show that the proposed method reduces the error as compared to the state-of-the-art for high-resolution images, and that the proposed scheme is more robust to blur in the imaged calibration pattern.
ContributorsPrakash, Charan Dudda (Author) / Karam, Lina J (Thesis advisor) / Frakes, David (Committee member) / Papandreou-Suppappola, Antonia (Committee member) / Arizona State University (Publisher)
Created2012
Description
Diabetic retinopathy (DR) is a common cause of blindness occurring due to prolonged presence of diabetes. The risk of developing DR or having the disease progress is increasing over time. Despite advances in diabetes care over the years, DR remains a vision-threatening complication and one of the leading causes of blindness among American adults. Recent studies have shown that diagnosis based on digital retinal imaging has potential benefits over traditional face-to-face evaluation. Yet there is a dearth of computer-based systems that can match the level of performance achieved by ophthalmologists. This thesis takes a fresh perspective in developing a computer-based system aimed at improving diagnosis of DR images. These images are categorized into three classes according to their severity level. The proposed approach explores effective methods to classify new images and retrieve clinically-relevant images from a database with prior diagnosis information associated with them. Retrieval provides a novel way to utilize the vast knowledge in the archives of previously-diagnosed DR images and thereby improve a clinician's performance while classification can safely reduce the burden on DR screening programs and possibly achieve higher detection accuracy than human experts. To solve the three-class retrieval and classification problem, the approach uses a multi-class multiple-instance medical image retrieval framework that makes use of spectrally tuned color correlogram and steerable Gaussian filter response features. The results show better retrieval and classification performances than prior-art methods and are also observed to be of clinical and visual relevance.
ContributorsChandakkar, Parag Shridhar (Author) / Li, Baoxin (Thesis advisor) / Turaga, Pavan (Committee member) / Frakes, David (Committee member) / Arizona State University (Publisher)
Created2012
Description
Communication amongst eusocial insect is key to their success. Ants rely on signaling to mediate many different functions within a colony such as policing and nest mate recognition. Camponotus floridanus uses chemosensory signaling in the form of cuticular hydrocarbons to regulate these functions. Each cuticular hydrocarbon profile contains numerous hydrocarbons, however it is yet to be seen if Camponotus floridanus can discriminate between linear hydrocarbons of similar length. Individual specimens were conditioned in three different ways: 5 conditioning with high concentration of sugar water (1;1 ratio), 1 conditioning with high concentration of sugar water, and 5 conditioning with low concentration of sugar water (1;4). Two linear hydrocarbons were use, C23 and C24, with C23 always being the conditioned stimulus. Specimens who were conditioned 5 times with high concentration of sugar water were the only group to show a significant response to the conditioned stimulus with a p-value of .008 and exhibited discrimination behavior 46% of the time. When compared 5 conditioning with high concentration to the other two testing conditioning groups, 1 conditioning with high concentration produced an insignificant p-value of .13 was obtained whereas when comparing it with 5 conditioning low concentration of sugar a significant p-value of .0132 was obtained. This indiciates that Camponotus floridanus are capable of discrimination however must be conditioned with high concentration of sugar water, while number of conditioning is insignificant.
ContributorsDamari, Ben Aviv (Author) / Liebig, Juergen (Thesis director) / Ghaninia, Majid (Committee member) / Pratt, Stephen (Committee member) / Barrett, The Honors College (Contributor) / School of Life Sciences (Contributor)
Created2014-05
Description
ABSTRACT Communication is vital in the context of everyday life for all organisms, but particularly so in social insects, such as Z. nevadensis. The overall lifestyle and need for altruistic acts of individuals within a colony depends primarily on intracolony chemical communication, with a focus on odorants. The perception of these odorants is made possible by the chemoreceptive functions of sensilla basiconica and sensilla trichoid which exist on the antennal structure. The present study consists of both a morphological analysis and electrophysiological experiment concerning sensilla basiconica. It attempts to characterize the function of neurons present in sensilla basiconica through single sensillum recordings and contributes to existing literature by determining if a social insect, such as the dampwood termite, is able to perceive a wide spectrum of odorants despite having significantly fewer olfactory receptors than most other social insect species. Results indicated that sensilla basiconica presence significantly out-paced that of sensilla trichoid and sensilla chaetica combined, on all flagellomeres. Analysis demonstrated significant responses to all general odorants and several cuticular hydrocarbons. Combined with the knowledge of fewer olfactory receptors present in this species and their lifestyle, results may indicate a positive association between the the social complexity of an insect's lifestyle and the number of ORs the individuals within that colony possess.
ContributorsMcGlone, Taylor (Author) / Liebig, Juergen (Thesis director) / Ghaninia, Majid (Committee member) / Barrett, The Honors College (Contributor)
Created2015-05
Description
This thesis describes the development, characterization, and application of new biomedical technologies developed around the photoacoustic effect. The photoacoustic effect is defined as optical absorption-based generation of ultrasound and provides the foundation for a unique method of imaging and molecular detection. The range of applications of the photoacoustic effect have not yet been fully explored. Photoacoustic endoscopy (PAE) has emerged as a minimally invasive tool for imaging internal organs and tissues. One of the main themes of this dissertation involves the first reported dual-intrauterine photoacoustic and ultrasound deep-tissue imaging endoscope. This device was designed to enable physicians at the point-of-care to better elucidate overall gynecological health, by imaging the lining of the human uterus. Intrauterine photoacoustic endoscopy is made possible due to the small diameter of the endoscope (3mm), which allows for complete, 360-degree organ analysis from within the uterine cavity. In certain biomedical applications, however, further minimization is necessary. Sufficiently small diameter endoscopes may allow for the possibility of applying PAE in new areas. To further miniaturize the diameter of our endoscopes, alternative imaging probe designs were investigated. The proposed PAE architecture utilizes a hollow optical waveguide to allow for concentric guiding of both light and sound. This enables imaging depths of up to several millimeters into animal tissue while maintaining an outer diameter of roughly 1mm. In the final focus of this dissertation, these waveguides are further investigated for use in micropipette electrodes, common in the field of single cell electrophysiology. Pulsed light is coupled with these electrodes providing real-time photoacoustic feedback, useful in navigation towards intended targets. Lastly, fluorescence can be generated and collected at the micropipette aperture by utilizing an intra-electrode tapered optical fiber. This allows for a targeted robotic approach to labeled neurons that is independent of microscopy.
ContributorsMiranda, Christopher (Author) / Smith, Barbara S. (Thesis advisor) / Kodibagkar, Vikram (Committee member) / LaBaer, Joshua (Committee member) / Frakes, David (Committee member) / Barkley, Joel (Committee member) / Arizona State University (Publisher)
Created2021