Kah Bin Lim

A prism-based single-lens stereovision system: From trinocular to multi-ocular

This paper investigates a passive stereovision system which employs one single CCD camera and one pyramid-like glass prism. Its trinocular variety is first presented: each image captured by this system can be split into three sub-images and these sub-images are taken as the images simultaneously captured by three virtual cameras which are generated by the prism. Two different approaches are developed to model this system: one bases on a conventional camera calibration technique and the other bases on geometrical analysis of ray sketching. The second approach is a relatively simpler but is sufficiently accurate as compared to the calibration based approach. Then the knowledge on this trinocular system is extended to build a single-lens multi-ocular stereovision system. Experiments are conducted to validate this system. The ideas presented in this paper are believed novel.

A Robotic System for Overlapping Radiofrequency Ablation in Large Tumor Treatment

Overlapping radiofrequency ablation (RFA) in large tumor treatment requires multiple insertions of electrodes, which often compromise its efficacy and predictability. Surgical robot is a promising candidate for the execution of multiple RFA in large tumor treatment in terms of accuracy and speed. In this paper, we share our experience of design and implementation of a novel robotic system specialized for overlapping RFA. It consists of two components: a robotic manipulator and an automatic ablation planning module. The manipulator architecture is designed to facilitate the kinematic requirement for the multiple overlapping ablation technique within the constraints of minimally invasive surgery. An efficient “Voxel Growing” algorithm is adopted to automatically produce the ablation points according to the tumors profile. The feasibility of the proposed robotic system is demonstrated by extensive simulation- and experiment-based evaluation conducted on ex vivo porcine liver.

Detection and estimation of circular arc segments

We present a series of algorithms for detecting circular arc segments and estimating their circular centers and radii from planar curves. Scale-space filtering is first used to remove noise and extract curve segments with similar local point curvatures. Arc extending and merging algorithms are then applied to extend and join these segments. Finally, perceptually significant circular arc segments are produced after the arc verification process.

A scale-space filtering approach for visual feature extraction

This paper presents a new integrated approach for detecting visual features which include CORNERs, ENDs, ARCs and LINEs. The effect of scale-space filtering on visual features is studied in detail as it forms the theoretical basis of our work. In this approach, the outline of the object is first extracted and it is then smoothed by scale-space filtering at different scale levels. Subsequently, the Local Extreme Curvature Points extracted from the smoothed curve and END candidates are determined to guide the termination of the filtering process. Information about the curvature of each point at the largest scale level is used to detect the different kinds of visual features. Several algorithms are proposed to determine CORNERs, ENDs, ARCs and LINEs. Experimental results show that our approach is robust to translation, rotation and scaling of the object as well as noise corruption. In addition, efficient visual features can also be successfully extracted with this approach.

A pragmatic 3D visual servoing system

This paper presents an attractive position-based visual servoing approach for camera-in-hand robotic systems. The major contribution of this work is in devising an elegant and pragmatic approach for 3D visual tracking. The proposed modified Smith predictor (MSP)-DeMenthon-Horaud (DH) visual servoing system has shown to be reliable and yielded good target tracking performance. It differ from the other techniques found in the vast visual servoing literature, in its approach for image interpretation and the introduction of a Smith-like predictor control structure to overcome the inherent vision delay. A complete description will be made on the MSP-DH visual servoing system. Experiments on the target tracking performance on AY planar motion using an AdeptOne robotic system are presented to illustrate the controller performance. Also, experimental results clearly showed the capability of the MSP-DH visual servoing system in performing 3D-dynamic visual servoing.

Comparison of the support vector machine and relevant vector machine in regression and classification problems

In this paper, we introduce the relevant vector machine (RVM) from Michael Tipping. The formulation of the RVM in regression and classification is reviewed. Then we analyze why the RVM can reach a sparse solution. In the experiment, we use the real application data to compare the performance of SVM and RVM. The advantages and disadvantage of the SVM and RVM is analyzed based on the experimental results. Some suggestion for the RVM is presented in the discussion section.

Intraoperative visual guidance and control interface for augmented reality robotic surgery

We proposed an augmented reality (AR) robotic system equipped with intraoperative visual guidance and gesture based control interface. The proposed feature is an enhancement of our AR robotic system. AR robotic system was introduced to the field of interventional medicine to assist surgeons in implementing medical operations under the augmented reality environment. The technique combining dextrous robot and AR guidance provides a new operational mode for surgeons. The introduction of robotic modules is to compliment surgeons dexterity and to perform specific tasks defined during the surgery. Augmented reality provides additional visualization and interaction absence in a typical clinical environment. The interfacing system includes intraoperative tracking of surgical tools and virtual reconstruction of the visually occluded tool segment, and an intuitive gesture-based human-computer interaction centered on the projected organ. Preliminary experiments show that this novel human-machine interface is effective for surgical intervention.

A new segment-based stereo matching using graph cuts

In the paper, we formulate a new energy function followed by the use of graph cuts to refine the disparity map which takes segment as node. Firstly, the robust disparity plane fitting is modeled and the method of Singular Value Decomposition (SVD) is used to solve least square. In order to ensure reliable pixel sets for the segment, we filter out outliers through three main rules, namely; cross-checking, judging reliable area and measuring the distance between previous disparity to the computed disparity plane. Secondly, we apply improve hierarchical clustering algorithm to merge neighbor. Finally, the final disparity map is obtained. Experimental results demonstrate that our approach is effective in improving the state of the art.

2D occluded object recognition using wavelets

A 2D object recognition algorithm applicable for partial occluded object recognition is proposed. The boundary of object of interest is extracted first. Then we segment the boundary into curve segments using dominant points, followed by a proportional extension. Normalization is then performed for each segment to make them translation, orientation and scaling invariant. After that, each segment is represented by its wavelet descriptors at multi-scale. A hierarchical iterative matching is performed to identify the object from low to high resolution. Experiment result shows proposed recognition algorithm is robust to similarity transform, noise and occlusion, and it is computational efficient.

Single-lens stereovision system using a prism: position estimation of a multi-ocular prism

In this paper, a position estimation method using a prism-based single-lens stereovision system is proposed. A multifaced prism was considered as a single optical system composed of few refractive planes. A transformation matrix which relates the coordinates of an object point to its coordinates on the image plane through the refraction of the prism was derived based on geometrical optics. A mathematical model which is able to denote the position of an arbitrary faces prism with only seven parameters is introduced. This model further extends the application of the single-lens stereovision system using a prism to other areas. Experimentation results are presented to prove the effectiveness and robustness of our proposed model.