Mei Yeen Choong

Image segmentation via normalised cuts and clustering algorithm

Image segmentation has been widely applied in image analysis for various areas such as biomedical imaging, intelligent transportation systems and satellite imaging. The main goal of image segmentation is to simplify an image into segments that have a strong correlation with objects in the real world. Homogeneous regions of an image are regions containing common characteristics and are grouped as single segment. One of the graph partitioning methods in image segmentation, normalised cuts, has been recognised producing reliable segmentation result. To date, normalised cuts in image segmentation of various sized images is still lacking of analysis of its performance. In this paper, segmentation on synthetic images and natural images are covered to study the performance and effect of different image complexity towards segmentation process. This study gives some research findings for effective image segmentation using graph partitioning method with computation cost reduced. Because of its cost expensive and it becomes unfavourable in performing image segmentation on high resolution image especially in online image retrieval systems. Thus, a graph-based image segmentation method done in multistage approach is introduced here.

Graph-Based Image Segmentation Using K-Means Clustering and Normalised Cuts

Image segmentation with low computational burden has been highly regarded as important goal for researchers. Various image segmentation methods are widely discussed and more noble segmentation methods are expected to be developed when there is rapid demand from the emerging machine vision field. One of the popular image segmentation methods is by using normalised cuts algorithm. It is unfavourable for a high resolution image to have its resolution reduced as high detail information is not fully made used when critical objects with weak edges is coarsened undesirably after its resolution reduced. Thus, a graph-based image segmentation method done in multistage manner is proposed here. In this paper, an experimental study based on the method is conducted. This study shows an alternative approach on the segmentation method using k-means clustering and normalised cuts in multistage manner.

Multistage Image Clustering and Segmentation with Normalised Cuts

Normalised cuts algorithm requires massive similarity measurement computation for image segmentation. Since a digital camera at present has the capability to produce high resolution image, it will be inevitably that resizing image into suitable resolution at which the algorithm can perform image segmentation with minimal burden. While retaining the important features in the images, natural images are likely to be restricted for resizing them into a particular smaller resolution. Dividing an image into equal size of regions (named as image cells) for the segmentation is proposed here to solve the problem of missing important features when the image resolution is overly reduced. Gradually, the locally segmented clusters from the image cells are taken for second stage segmentation to merge them up globally. In this paper, experimental results using the mentioned method are shown. Experiment shows that it is capable to produce reasonable segmented clusters based on the proposed approach.

Enhancement of Particle Filter Resampling in Vehicle Tracking Via Genetic Algorithm

Vehicle tracking is an essential approach that can help to improve the traffic surveillance or assist the road traffic control. Recently, the development of video surveillance infrastructure has incited the researchers to focus on the vehicle tracking by using video sensors. However, the amount of the on-road vehicle has been increased dramatically and hence the congestion of the traffic has made the occlusion scene become a challenge task for video sensor based tracking. Conventional particle filter will encounter tracking error during and after occlusion. Besides that, it also required more iteration to continuously track the vehicle after occlusion. Thus, particle filter with genetic operator resampling has been proposed as the tracking algorithm to faster converge and keep track on the target vehicle under various occlusion incidents. The experimental results show that enhancement of the particle filter with genetic algorithm manage to reduce the particle sample size.

Trajectory Clustering for Behavioral Pattern Learning in Transportation Surveillance

The development of an efficient traffic flow monitoring system has been the main focus for many researchers working in the field. Due to the rapid development in urbanization, the complexity of traffic intersections provides challenges for researchers to detect the underlying traffic scenes. With the emerging video based surveillance system, vehicle trajectory can be extracted for observation and prediction via behavioral pattern learning. Prior to the learning, clustering of the extracted vehicle trajectory data is performed to group the data based on similarity measures. In this paper, the implementation of clustering algorithm on the trajectory data is analyzed and issues concerning the trajectory clustering are discussed.

Clustering Algorithm in Normalised Cuts Based Image Segmentation

Normalised cut method has been effectively used for image segmentation by representing an image as weighted graph in global view. It does segmentation via partitioning the graphs into sub-graphs. Clustering algorithm is implemented such that sub-graphs with common similarities are grouped together into one cluster and separates sub-graphs that are dissimilar into distinctive clusters. Clustered segments from the normalised cuts are then produced. As the clusters initialisation gives influence to the segmentation result, optimisation of the clustering algorithm is implemented to achieve better segmentation. With the approach applied in the normalised cuts based image segmentation, the constraint of using normalised cuts algorithm in image segmentation can be alleviated. In this paper, evaluation of the clustering algorithm with the normalised cuts image segmentation on images has been carried out and the effect of different image complexity towards normalised cuts segmentation process is presented.

Foetus Ultrasound Medical Image Segmentation via Variational Level Set Algorithm

There is a challenge to segment the medical image which is often blurred and consists of noise. The objects to be segmented are always changing shape. Thus, there is a need to apply a method to automated segment well the objects for future analysis without any assumptions about the objects topology are made. In general, when performing pregnancy ultrasound scanning, obstetrician needs to find out the best position or angle of the foetus and freeze the scene. The obstetrician will click on the crown and the rump of the foetus to get the foetus length. The segmentation technique applied is level set method. A variational level set algorithm has been successfully implemented in medical image segmentation (Xray image, MRI image and ultrasound image). The results showed the level set contour evolved well on the low contrast and noise consisting medical image, especially the ultrasound image.

Implementation of image segmentation on foetus ultrasound imaging system

Obstetrics ultrasound scan has been a vital routine for a pregnant mother to get information on the foetus dating and growth. Foetus ultrasound image is normally not clear and contains unwanted noise. Furthermore, the displayed foetus scan on the monitor screen can be not in complete stationary because of the slight movement of the held ultrasound probe. Thus, a computerized method to do segmentation on the foetus image should be implemented. To obtain precise measurements, obstetrician needs to freeze the best possible scene throughout the scanning session. With the segmentation technique implemented, the point locations for measurement can be generated without the participation of the obstetrician. In this paper, the applied segmentation technique is variational level set algorithm. Based on the segmentation results, the level set contour evolved well on the ultrasound image although it is low in contrast and contains image noise.