Yuille et al  used global information of the eye to improve the extraction process. Once established near an eye feature, optimal feature boundaries are minimized using steepest gradient descent minimization. Its Limitations is that it is sensitive to initial placement and the processing time is high. Smart Snakes or Point Distributed Models (PDMs) are compact parameterized descriptions of a shape based upon statistics. They use PCA to construct a linear flexible model from variations of the features in a training set.
There are many approaches to find accurate estimation of Gaze. There is a popular technique to solve this problem is electro-oculography -.In this technique the sensors are used i.e electrodes. sensor is for collecting the information about position of eyeball. Another approach is first track the face and then detect and track the eyes .There is Feature based and Image based method to detect he face. In the feature based method, finding the facial feature such as nose, eyes brows, eyes pupil and lips.
involved in iris recognition process are given as follow. (i) Image acquisition- It interfaces between the real world and the system; it has to acquire all the necessary data. In this step, a high – quality image of an eye is captured. (ii) Iris segmentation and location- Sometimes necessary pre-processing has to be done to remove artifacts, to enhance (e.g. removing background noise) the image.
The image stitching can be divided into three main steps: image calibration, image registration, and image blending , as shown in figure 1. Image calibration produces an estimate of the intrinsic and extrinsic camera parameters. Inimage registration, multiple images are compared to find the translations that can be used for the alignment of images. After registration, these images are merged together to form a single image. In the following subsections, these main steps are discussed briefly.
Viet Dzung Nguyen et al., proposed a detection process based on local contrast thresholding and rule-based classification which was performed over the preprocessed and segmented mammograms . J.C. Nunes et al., presented a texture analysis algorithm based on Gray-Level Cooccurrence (GLC) model and Bidimensional Empirical Mode Decomposition (BEMD) of a texture field . Gabriel Rilling et al., presented, Huang’s data-driven technique of Empirical Mode Decomposition (EMD), and issues related to its effective implementation were
The eigen vectors and eigen values of the covariance matrix are calculated. Now we will be having the feature vectors. Voila jones , the algorithm simply performs an exhaustive search using a sliding window, using different sizes, aspect ratios, and locations. The block diagram is as shown in Figure 1 Method for eigen face technique consists of the following
 used segmentation to group the image into regions with same property or characteristics. Methods of image segmentation include simple thresholding, K-means Algorithm and Fuzzy C-means. Microaneurysms are distinguished from nonmicroaneurysms by a set of decision criteria which includes pixel counts, ratio of minor to major axes, robustness test, length test and holes test. The region within the range of MinPixelCount is treated as MA. A region less than the range ofMinPixelCount is treated as a background noise while a region greater than the range of MinPixelCount is treated as a non-microaneurysm.
Abstract Remote-sensing classification is a complex process and requires consideration of many factors. This paper investigates the process of remote sensing classification. Importance is placed on the summarization of major advanced classification approaches and the techniques used for improving classification accuracy. This paper suggests that designing a suitable image processing procedure is a prerequisite for a successful classification of remotely sensed data into a thematic map. The major steps of image classification may include determination of a suitable classification system, selection of training samples, image pre-processing, feature extraction, selection of suitable classification approaches, post-classification processing, and
ABSTRACT: Image mining is a part of data mining in the field that handles with the hidden knowledge extraction. Scanned images plays a significant role in image mining research, it analysis the images using numerous techniques which improves the quality of the images. The two important components of image mining are image enhancement and information extraction. Image enhancement techniques are used to improve the quality of image where as information extraction techniques extract the statistical information about portion of an image or any particular image. Document images may contain both scanned machine printed documents and handwritten documents.