This paper presents the comparative study of face recognition using discrete orthogonal moment namely Krawtchouk moments (KMs) and Tchebichef moments (TMs). Both these moments do not require numerical approximation an...
详细信息
This paper presents the comparative study of face recognition using discrete orthogonal moment namely Krawtchouk moments (KMs) and Tchebichef moments (TMs). Both these moments do not require numerical approximation and coordinate space normalization. The complex computation of radial polynomial as order becomes larger is not an issue and this makes KMs and TMs superior compared to continuous orthogonal moments in terms of preserving the analytical properties needed to ensure minimal information redundancy. With these properties, KMs and TMs are well suited as pattern features in the analysis of two-dimensional images. The selection of orders and parameters of KMs determines the ROI and to obtain the reconstructed face image similar to the original face image, both parameters of KMs are set equal at 0.5. As for TMs, only the selection of order is required. The experiments were carried out on the database face images from the AT&T Laboratories Cambridge University consisting of 40 distinct subjects of 10 non-similar images each. For some subjects, the images were taken at different times, varying the lighting, facial expressions (open/closed eyes, smiling/not smiling), facial details (with and without spectacles) and different face scale. Euclidean square distance or Nearest Neighbour (NN) is used as the classifier in the recognition stage. From the experiments, KMs showed better performance as compared to TMs in terms of classification accuracy.
As an important tool to study practical problems of biology, engineering and imageprocessing, the neuralnetworks has caused more and more attention. Some interesting results on the stability have been obtained. In t...
详细信息
ISBN:
(纸本)9780769538167
As an important tool to study practical problems of biology, engineering and imageprocessing, the neuralnetworks has caused more and more attention. Some interesting results on the stability have been obtained. In this paper, the exponential stability of the equilibrium point of a group of Cohen-Grossberg neuralnetworks is obtained by using Lyapunov method and Razumikhin technique.
The main objective of this project is to develop a technique to classify the ripeness of bananas into 3 categories, which is unripe, ripe and overripe systematically based on their histogram RGB value components. This...
详细信息
ISBN:
(纸本)9781424455614
The main objective of this project is to develop a technique to classify the ripeness of bananas into 3 categories, which is unripe, ripe and overripe systematically based on their histogram RGB value components. This system involved the process of collecting samples with different level of ripeness, imageprocessing and image classification by using artificialneural network. Collecting bananas sample is done by using Microsoft NX6000 webcam with 2 mega pixels. 32 samples were used as training samples for artificialneural network. In order to see whether the method mention above can classify the image correctly, another 28 images was used as a testing. From the result obtained, it was shown that the artificialneural network can generally classify the ripeness of bananas. This is because it can classify up to 25 samples correctly out of 28 samples. Developing a program totally by using Matlab version 7.0 can help classification process successfully.
For the automatic inspection for printed labels, which are covered with rubber-like coatings and Curl, we have developed a camera-based portable inspection system. In this paper, we explained the developed system, and...
详细信息
ISBN:
(纸本)9783642024801
For the automatic inspection for printed labels, which are covered with rubber-like coatings and Curl, we have developed a camera-based portable inspection system. In this paper, we explained the developed system, and especially discuss the inspection method of the spread and chip of the printed labels using neuralnetworks. The experimental results confirm the validity of the proposed method for the spread and chip of alphanumerics.
In this paper we present a hardware architecture for a Support Vector Machine intended for vision applications to be implemented in a FPGA device. The architecture computes the contribution of each support vector in p...
详细信息
ISBN:
(纸本)9783642042737
In this paper we present a hardware architecture for a Support Vector Machine intended for vision applications to be implemented in a FPGA device. The architecture computes the contribution of each support vector in parallel without performing multiplications by using a CORDIC algorithm and a hardware-friendly kernel function. Additionally input images are not preprocessed for feature extraction as each image is treated as a point in a high dimensional space.
The vision has many sensors responsible for capturing information that is sent to the brain. The gaze reflects its attention, intention and interest of the brain towards the outside world. Therefore, the detection of ...
详细信息
ISBN:
(纸本)9781424435494
The vision has many sensors responsible for capturing information that is sent to the brain. The gaze reflects its attention, intention and interest of the brain towards the outside world. Therefore, the detection of the gaze direction is a promising alternative for the simulation programs, virtual reality applications and human machine special communication. Cheaper devices to capture images and increase the power processing of personal computers motivate studies that allow human-machine interactivity. The application of techniques to detect the gaze direction has the possibility of improving significantly the interaction between people with motor deficiency and personal computers. The objective of this work is to provide a system that uses techniques of digital imageprocessing to classify the gaze direction. The results show the complexity and efficiency of a system that performs not only the acquisition of images but also their classification by using artificialneuralnetworks.
Dynamic knowledge increase of associative memory is essential for practical applications of artificialneuralnetworks. However existing discrete bipolar neuralnetworks have no properties to achieve this aim. In this...
详细信息
ISBN:
(纸本)9780769536347
Dynamic knowledge increase of associative memory is essential for practical applications of artificialneuralnetworks. However existing discrete bipolar neuralnetworks have no properties to achieve this aim. In this paper, we proposed an new multi module associative memory model for many to many associations based on incidence of patterns, through adding new neurons and connections to neuralnetworks of this model, which may increase knowledge dynamically as well as not forget information stored before. The properties of dynamic knowledge increase in new associative memory are investigated in detail.
This paper introduces the application of the feature transformation approach proposed by Torkkola [1] to the domain of imageprocessing. Thereto, we extended the approach and identifed its advantages and limitations. ...
详细信息
ISBN:
(纸本)9783642042768
This paper introduces the application of the feature transformation approach proposed by Torkkola [1] to the domain of imageprocessing. Thereto, we extended the approach and identifed its advantages and limitations. We compare the results with more common transformation methods like Principal Component Analysis and Linear Discriminant Analysis for a function approximation task from the challenging domain of video-based combustion optimization. It is demonstrated that the proposed method generates superior results in very low dimensional subspaces. Further, we investigate the usefulness of an adaptive variant of the introduced method in comparison to basic subspace transformations and discuss the results.
In this paper, some of the imageprocessing and pattern recognition methods that have been used on medical images for cancer diagnosis are reviewed. Previous studies on artificialneuralnetworks, Genetic Programming,...
详细信息
ISBN:
(纸本)9781424453306
In this paper, some of the imageprocessing and pattern recognition methods that have been used on medical images for cancer diagnosis are reviewed. Previous studies on artificialneuralnetworks, Genetic Programming, and Wavelet Analysis are described with their working process and advantages. The definition of each method is provided in this study, and the acknowledgement is granted for previous related research activities.
An automatic power quality disturbance measurement system has been proposed in this paper. The probabilistic entropy is calculated for the most frequently occurring short duration disturbances in distribution system w...
详细信息
ISBN:
(纸本)9781424455614
An automatic power quality disturbance measurement system has been proposed in this paper. The probabilistic entropy is calculated for the most frequently occurring short duration disturbances in distribution system which is based on Shannon's entropy. The variation of probabilistic entropy with respect to the deviation in the magnitude and duration of the disturbed signal is presented in the form of bar chart and graphs. The automatic recognition is achieved using an artificialneural network, (ANN) in conjunction with fuzzy based decision logic. A new type of power quality (PQ) factor, Equivalent Disturbance Factor (EDF) has been proposed in this paper. Then ANN has been implemented to predict the Equivalent Disturbance Factor.
暂无评论