In this paper an algorithm to cluster face images found in video sequences is proposed. A novel method for creating a dissimilarity matrix using SIFT image features is introduced. This dissimilarity matrix is used as ...
详细信息
ISBN:
(纸本)9781424407071
In this paper an algorithm to cluster face images found in video sequences is proposed. A novel method for creating a dissimilarity matrix using SIFT image features is introduced. This dissimilarity matrix is used as an input in a hierarchical average linkage clustering algorithm, which yields the clustering result. Three well known clustering validity measures are provided to asses the quality of the resulting clustering, namely the F measure, the overall entropy (OE) and the Gamma statistic. The final result is found to be quite robust to significant scale, pose and illumination variations, encountered in facial images.
Discrete signalprocessing using fuzzy fractal dimension and grade of fractality is proposed based on the novel concept of merging fuzzy theory and fractal theory. The fuzzy concept of fractality, or self-similarity, ...
详细信息
ISBN:
(纸本)9781424407071
Discrete signalprocessing using fuzzy fractal dimension and grade of fractality is proposed based on the novel concept of merging fuzzy theory and fractal theory. The fuzzy concept of fractality, or self-similarity, in discrete time series can be reconstructed as a fuzzy-attribution, i.e., a kind of fuzzy set. The objective short time series can be interpreted as an objective vector, which can be used by a newly proposed membership function. Sliding measurement using the local fuzzy fractal dimension (LFFD) and the local grade of fractality (LGF) is proposed and applied to fluctuations in seawater temperature around the Izu peninsula of Japan. Several remarkable characteristics are revealed through "fuzzy signalprocessing" using LFFD and LGF.
In this paper, a novel color image quantization algorithm is presented. This new algorithm addresses the question of how to incorporate the principle of human visual perception to color variation sensitivity into colo...
详细信息
ISBN:
(纸本)9781424407071
In this paper, a novel color image quantization algorithm is presented. This new algorithm addresses the question of how to incorporate the principle of human visual perception to color variation sensitivity into color image quantization process. Color variation measure (CVM) is calculated first in CIE Lab color space. CVM is used to evaluate color variation and to coarsely segment the image. Considering both color variation and homogeneity of the image, the number of colors that should be used for each segmented region can be determined. Finally, CF-tree algorithm is applied to classify pixels into their corresponding palette colors. The quantized error of our proposed algorithm is small due to the combination of human visual perception and color variation. Experimental results reveal the superiority of the proposed approach in solving the color image quantization problem.
This paper examines the feasibility of an approach to image retrieval from a heterogeneous collection based on texture. For each texture of interest (T), a T-vs-other classifier is evolved for small n x n windows usin...
详细信息
ISBN:
(纸本)9781424407071
This paper examines the feasibility of an approach to image retrieval from a heterogeneous collection based on texture. For each texture of interest (T), a T-vs-other classifier is evolved for small n x n windows using genetic programming. The classifier is then used to segment the images in the collection. If there is a significant contiguous area of T in an image, it is considered to contain that texture for retrieval purposes. We have experimented with sky and grass textures in the Corel Volume 12 image set Experiments with a single image indicate that classifiers for the two textures can be learned to a high accuracy. Experiments with a test set of 714 Corel images gave a retrieval accuracy of 84% for both sky and grass textures. These results suggest that the use of texture could enhance retrieval accuracy in content based image retrieval systems.
In this paper a method for image segmentation using an opposition-based reinforcement learning scheme is introduced. We use this agent-based approach to optimally find the appropriate local values and segment the obje...
详细信息
ISBN:
(纸本)9781424407071
In this paper a method for image segmentation using an opposition-based reinforcement learning scheme is introduced. We use this agent-based approach to optimally find the appropriate local values and segment the object. The agent uses an image and its manually segmented version and takes some actions to change the environment (the quality of segmented image). The agent is provided with a scalar reinforcement signal as reward/punishment. The agent uses this information to explore/exploit the solution space. The values obtained can be used as valuable knowledge to fill the Q-matrix. The results demonstrate potential for applying this new method in the field of medical image segmentation.
In classical graph-based image segmentation, a data-driven matrix is constructed representing similarities between every pair of pixels. The eigenvectors of such matrices contain relevant information about the cluster...
详细信息
ISBN:
(纸本)9781424407071
In classical graph-based image segmentation, a data-driven matrix is constructed representing similarities between every pair of pixels. The eigenvectors of such matrices contain relevant information about the clusters present on the image. An approach to image segmentation using spectral clustering with out-of-sample extensions is presented. This approach is based on the weighted kernel PCA framework. An advantage of the proposed method is the possibility to train and validate the clustering model on subsampled parts of the image to be segmented. The cluster indicators for the remaining pixels can then be inferred using the out-of-sample extension. This subsampling scheme can be used to reduce the computation time of the segmentation. Simulation results with grayscale and color images show improvements in terms of computation times together with visually appealing clusters.
The segmentation of MR images has been playing an important role to improve the detection and diagnosis of breast cancer. Main problem in breast images is the identification of the boundary between chest wall and brea...
详细信息
ISBN:
(纸本)9781424407071
The segmentation of MR images has been playing an important role to improve the detection and diagnosis of breast cancer. Main problem in breast images is the identification of the boundary between chest wall and breast tissue. Minimizing the effects of patient motion is also important step in segmentation process. In imageprocessing, there are many different segmentation algorithms. The most common used method among them is thresholding. However, classic thresholding methods are not effective for axial MR breast images completely because of the fact that the sequence artifacts in axial MR breast images are very high. For this reason, we have proposed a regional thresholding algorithm to segment MR images successfully. The outstanding problem is how to obtain an automatic procedure for detecting boundary between breast tissue and chest wall.
This paper introduces a novel approach to change gathered images from WWW into training images to build an image thesaurus. The requirements for being training images are a large number of images and with highly relev...
详细信息
ISBN:
(纸本)9781424407071
This paper introduces a novel approach to change gathered images from WWW into training images to build an image thesaurus. The requirements for being training images are a large number of images and with highly relevant to a given concept. To fulfill these requirements, a system should be able to collect a large number of relevant images to a given concept from WWW by the proposed criterion of relevance to the concept for each image. Then, the irrelevant images would be filtered out by the modified hierarchical clustering method based on the weighted combination of 5 MPEG-7 visual descriptors[9] and the proposed criterion of relevance to the concept for each cluster. Upon experimental results, the precision of the set of images generated by the proposed method is about 18% higher than that of the set of images generated by other methods[1][2].
A machine vision based keg inspection system can allow cost effective keg tracking and preventative maintenance programs to be implemented, leading to substantial savings for breweries with large keg fleets. A robust ...
详细信息
ISBN:
(纸本)9781424407071
A machine vision based keg inspection system can allow cost effective keg tracking and preventative maintenance programs to be implemented, leading to substantial savings for breweries with large keg fleets. A robust keg serial number recognition and keg condition assessment process is required to cater for different keg brands and a range of keg ages in the fleet It has been demonstrated that the proposed imageprocessing methodology, and neural network based number recognition system, successfully located and identified keg serial numbers with a 92% digit accuracy. Furthermore, the vision system allowed the concurrent assessment of the keg condition by assessing deformity of the keg rim, and that of the filler valve. A correlation coefficient, generated using a template matching process, proved to be a suitable metric which adequately indicated rims within and outside acceptable deformity bounds.
暂无评论