No self-introduction
Loading
Profile
Follow
codes (7)
Gabor Feature Extraction For Emotion Detection of
no vote
A Gabor filtering method to describe texture features is proposed, which is suitable for the probability wave probability transformation of physical properties and the calculation of texture features; 2) adaptive strategy and feature weighting of function and filter selection is proposed, in order to reduce the feedback in redundant representation and retrieve the information needs of users in the image by using the correlation of users; 3) integration of the two plans Content based breast retrieval recommendation based on Gabor filtering method and adaptive strategy
vicky948285 0 0
1
SOM for NN Training
5.0
An important aspect of the artificial neural network model is whether it needs to be learned or not. Based on their learning, all the way artificial neural networks can be divided into two categories of learning - supervision and supervision. In supervised learning, the network is trained according to the output result of each input, and the output result is needed when the vector is used. Supervised learning types of artificial neural networks, such as multilayer perceptrons, use the neural parameters to guide the formation of target results. It is therefore possible for neural networks to learn behavior according to the research process. In unsupervised learning, the training of the network is completely data-driven and no target input is provided for the data vector results. Types of unsupervised learning neural networks, such as self-organizing maps, can be used to cluster the input data and find the inherent characteristics of the problem.
vicky948285 1 0
1
Geometric features
no vote
Before building a face detection system, it is very useful if we know how eyes function. In general, eyes can detect a wide variety of objects under various environment. However, if an object appears at a particular pose, e.g., upside down, we may not recognize it at once. The reason is the orientation of this object is different from that of its counterpart in memory. Therefore, when one wants to check what exists in an image, usually the whole geometrical configuration is checked instead of checking object's details. Besides, since a face contains four main organs, i.e., eyebrows, eyes, nose, and mouth, it is very important to detect these facial features so that the whole geometrical configuration of the face can be identified. In general, eyebrows, eyes, nostrils and mouth always look darker than the rest of a face due to illumination. Among them, the darkness of eyebrows also depends on their density and color, and the nostrils may not always be seen.
vicky948285 0 0
1
face recognition
no vote
This code has been designed for facial detection using matlab software. Anyone can use this code for facial detection in biometric scanning system which will be useful for security purpose.
vicky948285 0 0
1
Automati road detection
no vote
he detail). The environment of intelligent vehicle is changing continuously while the vehicle is moving. However, because the process of classifier parameters computing is time and computation exhausting, it is not suitable to compute the parameters in every frame. Therefore, given the assumption that the environment does not change drastically in a few consecutive frames, we take the classifier parameters computing as a parallel process with all other components. From the experiments, the values of parameters are updated in every 8-12 frames. D. SVM Classifier There are two stages in the component of SVM Classifier: road detection classifier training and road detection classifier classification. 1) Road detection classifier training: Fig.4 gives an outline of the road detection classifier training stage. As we mentioned above, we use the SVM with the RBF kernel as the road detection classifier. G
vicky948285 0 0
1
feature extraction detection
no vote
In the online learning operation, our algorithm process acquires the interest points of eigenvectors. In view of the assumption that the road region is simply connected, the point is listed as a road, lying in the non road region, it can be seen that as a point of interest, it is marked as a negative sample (non road sample), and vice versa. We annotate these points as new training data (Figure 9). In the process of online learning practice, one thing we should pay attention to is that we don't know where the real boundary is. What we can get is the result of morphological operation on the edge of the road region. In order to reduce the possibility of mislabeling training data near the road edge, the threshold m is set to the width near the boundary of the margin, and the operation results of the road and non road areas are shown in FIG. 10. In our experiment, we set M to be 40 pixels wide
vicky948285 0 0
1
Image Denoising
no vote
An off-line handwritten alphabetical character recognition system using multilayer feed forward neural network is described in the paper. A new method, called, diagonal based feature extraction is introduced for extracting the features of the handwritten alphabets. Fifty data sets, each containing 26 alphabets written by various people, are used for training the neural network and 570 different handwritten alphabetical characters are used for testing. The proposed recognition system performs quite well yielding higher levels of recognition accuracy compared to the systems employing the conventional horizontal and vertical methods of feature extraction. This system will be suitable for converting handwritten documents into structural text form and recognizing handwritten names Handwriting recognition has been one of the most fascinating and challenging research areas in field of image processing and pattern recognition in the recent years [1
vicky948285 1 0
1
No more~
Follows0
Fans0