Feature Analysis Techniques

959 Words4 Pages

Abstract –Feature key point extraction plays a crucial role in the area of image processing applications (e.g. Medical). Before getting a feature, a variety of image pre processing techniques like thresholding the image, modifying the image size, ranging the pixel intensity values and binary operations are applied on the original image for noise removal process. After, getting features of images that will be useful in classification and object recognition purpose. Feature extraction is important to define the behavior of images; they show its place in important terms of storage taken in memory and efficiency in classification. This paper presents the method (SIFT) Scale Invariant Feature Transform detect and extract the local variations based …show more content…

The main aim of the extraction process will provide a "feature description" of an image. Feature description extracted from training image. Feature set containing the unique points of the image part which means the larger part of the image has been made in the form of detectable format. This can easily detectable and identifying the objects, boundaries and edges. Another important characteristic is to detect the larger number of features from the description of image sets which will reduces the redundant errors caused through local image variations of all matching feature point in an image. In the same way image morphological features are used to detect the internal and external property of images. This paper concludes the object recognition method and techniques for extraction of image information. The transformation mainly locate the directional and edge based approaches for finding the image information in the sense feature key points and the key points are mainly used for image classification …show more content…

General level of image features such as extraction of texture, shape, and color. This paper presents the concurrence matrix method to extract motion estimation image texture features namely; Inverse Difference Moment, Second Moment based of angular concepts with Correlation and Entropy values are calculated by using feature extraction techniques. A. Amali Asha and S.P. Vector et.al, explained the Learing algorithms for pattern recognition. Edge detection and object boundary detection concepts are discussed. Two different concepts are presented which include ant colony with gradient based existing edge detectors. C.I. Christodoulou and C.S. Pattichis described the multiple features from carotid plaque ultrasound images Categorization done on the plaque identification. And then morphology and structure based information retrieval process done based on the selection feature set. Finally neural network and statistical map were used for classification purpose. III. FEATURE EXTARCTION

More about Feature Analysis Techniques

Open Document