Motion-based Image Segmentation Research Articles

Modeling a Virtual Bare-Hand Interface System Using a Robust Hand Detection Approach for HCI

A practically deployable gesture recognition system is developed using a robust hand detection method implemented using a motion-based image segmentation process and a two-level bare hand classification model, which is integrated with a gesture classification system of 58 gestures using new robust features. Since detection of bare hand is affected by nonideal conditions, multiple color-texture features are analyzed in this study. In the second stage of the system, 18 new ASCII characters are introduced and analyzed along with the existing 40 characters (alphabets, numbers, and arithmetic operators). New 15 dimensional features are introduced along with the existing features to enhance the classification accuracy of the gestures. Significance of features statistically tested using one-way analysis of variance (ANOVA), Kruskal–Wallis and Friedman test, which are sequentially ranked and evaluated using incremental feature selection (IFS) method. Performance of the proposed hand detection system is observed to be 12.5% higher than the existing hand detection system under clean conditions, while 46.4% higher under the nonideal conditions. Performance of 58 gestures classification model has improved by 12.08% (Naïve Bayes), 8.86% (ELM), 10.83% (SVM), 8.02% ([Formula: see text]NN), and 6.61% (ANN) after using the new features. Majority voting-based classifier fusion method further improves the performance of the gesture recognition system by 3.88%, which is validated by Turkey’s HSD test.

Comparative Analysis of Motion base Image Segmentation using Machine Learning Techniques

Objectives: In the computer vision task, motion based image segmentation is the foremost step which warrants for further research. Several progressive techniques are being created for motion base image segmentation. Machine learning methods are implemented and analyzed in this paper. Methods: Motion is one of the useful characteristics to segment object from the image as it provides a classification of pixel in motion or motionless type. Machine learning methods are giving promising result in classification type problems. Machine learning methods reduce the execution time. In this type of technique a training is required in beginning after that all processes are performed automatically. This paper presents two machine learning technique SVM and LS-SVM. The Segmentation of image can be done by selecting image feature. In proposed method intensity change basis of motion is used as texture feature. Findings: Both of techniques classify pixels of an image between two types, 1. Appropriate to a motion and 2. Not appropriate to motion. Results obtained using this approach shows that machine learning methods are very promising techniques for this type of area. Application/Improvement: LS-SVM method is very promising with enhancement in execution time and segmented object result quality as compared to SVM.

THE BRAIN'S VIEW OF THE NATURAL WORLD IN MOTION: COMPUTING STRUCTURE FROM FUNCTION USING DIRECTIONAL FOURIER TRANSFORMATIONS

Attractive possibilities exist for the brain to solve important problems of visual processing in Fourier space. First, we describe a wave-interference model for computing motion contrast directly from the moving intensity distribution, without the need for directionally selective motion sensors. We then propose a global method for motion-based image segmentation based on unsupervised clustering in a three-dimensional Fourier space. The Fourier components of coherently moving entities are segregated from the remainder by means of a simple velocity proximity measure. This is accomplished without altering the spatial frequency components encoding the object, thereby ensuring that the spatiotemporal features of the segregated object can be reconstructed.

A look at... shape and function from motion in medical imaging: Part I

Image sequence analysis has been an active research area for many years. It brought new insights into a lot of distinct applications, ranging from site securing (intrusion detection) to image coding and video indexing as well as path determination or object handling in robotics. The pioneering contributions provided in the 1980s (e.g., the works of Hildreth, Nagel, and Horn and Schunck) aimed at two-dimensional (2-D) image sequence analysis through motion field estimation have benefited from major technological and methodological advances. New capabilities in capturing three-dimensional scenes have opened the road for emerging applications and have therefore led to more demanding solutions for object feature extraction, reconstruction, and interpretation. This column revisits some up-to-date problems in digital imaging. A future column will point out some of the challenges that should be considered in medical imaging. Topics covered include motion estimation; motion based image segmentation; and tracking.

No role for motion blur in either motion detection or motion-based image segmentation

The human contrast sensitivity function is bandpass in form for stimuli of low temporal frequency but low pass for flickering or moving stimuli. Because the loss in sensitivity to moving stimuli is large, images moving on the retina have little perceptible high-spatial-frequency content. The loss of high-spatial-frequency content--often referred to as motion blur--provides a potential cue to motion. The amount of motion blur is a function of stimulus velocity but is significant at velocities encountered by the visual system in everyday situations. Our experiments determined the influence of high-spatial-frequency losses induced by motion of this order on motion detection and on motion-based image segmentation. Motion detection and motion-based segmentation tasks were performed with either spectrally low-pass or spectrally broadband stimuli. Performance on these tasks was compared with a condition having no motion but in which form differences mimicked the perceptual loss of high spatial frequencies produced by motion. This allowed the relative salience of motion and motion-induced blur to be determined. Neither image segmentation nor motion detection was sensitive to the high-spatial-frequency content of the stimuli. Thus the change in perceptual form produced in moving stimuli is not normally used as a cue either for motion detection or for motion-based image segmentation in ordinary situations.

Contour from motion processing occurs in primary visual cortex.

Relative motion is one of the most salient cues for segmentation of a visual scene into separate objects. This is illustrated by the vivid contours that are perceived when random dot patterns move in different directions. Once motion is halted in such displays the segmentation contours disappear. This makes random dot patterns ideal for the study of contour from motion processing in isolation. Contour from motion processing obviously relies on direction-selective neurons, which are found in many visual cortical areas. It is, however, largely unknown at what level of processing their signals interact to serve the global process of motion-based image segmentation. To answer this question, we recorded visually evoked potentials, both in man and in awake monkey, to a stimulus specifically designed to signal the presence of neuronal activity related to contour from motion processing. We report here that response components specific to contour from motion were elicited only when the stimulus yielded a contour percept. In awake monkey, the sources of these components were located within the supra- and infra-granular layers of primary visual cortex. We conclude that V1 is involved in image segmentation processing.