International Journal of Information Acquisition

We present the two-stage adaptive metric learning procedure with improved generalization of missing training data for facial signature authentication. The conventional learning models suffer from degraded recognition rates due to poor estimation of decision boundary to classify impostor patterns. The two-stage networks combine multiple image synthesis methods to assume mixture patterns of training classes and facilitate threshold setting of false acceptance/rejection rates. The margin size of a decision boundary is adjusted to input patterns obtained from synthesized images and depends on the choice of the mixing factor. The present method effectively reduces the margin of a class with an improved recognition rate to classify impostor patterns from 82.2% to 98.8%. Furthermore, we examine the margin structure of the mixture distributions by using the Support Vector Machines.

Real-time force and visual information during MWCNT manipulation is required for online controlling MWCNT assembly based on atomic force microscope (AFM). Here real-time three-dimensional (3D) interactive forces between probe and sample are obtained according to PSD signals based on the proposed force model, and MWCNT manipulation process can be online displayed on the visual interface according to probe's position and applied force based on the proposed MWCNT motion model. With real-time force and visual information acquisition and feedback, the operator can control online MWCNT's manipulation process by adjusting the probe's 3D motion and applied forces. MWCNT push and assembly experiments verify the effectiveness of the method, which will be used in assembling MWCNT based nano device.

To assist the diagnosis, this paper presents a new algorithm to enhance the capsule endoscopy images, which are now widely used to detect the diseases in the digestive tract because of its great advantages over traditional endoscopy technologies. The capsule endoscopy images have the characteristics such as low resolution and low contrast, which cause some difficulties to analysis and diagnosis for the physicians, especially to computer aided diagnosis. To overcome these shortcomings, we propose a contrast-based forward and backward anisotropic diffusion method to enhance the capsule endoscopy images. Based on the local analysis of the capsule endoscopy images, we put forward a new concept of contrast. Then, we employ the forward and backward anisotropic diffusion to enhance the capsule endoscopy images. Experimental results show that this new method can efficiently enhance the capsule endoscopy images, which provides a better visualization and an improved discrimination of region of interest from the background.

One of the challenging problems of 3D Euclidean reconstruction is that only the hypothesis of intrinsic parameters can be used to retrieve the camera parameters. In this paper, we proposed a method to find out the intrinsic parameters of a camera using the rank constraint of the relation matrix of absolute conic Ω. Simultaneously, analyzed the degeneration of self-calibration. The experimental results showed that the self-calibration method of camera is better than the other methods.

The simultaneous localization and mapping technique is an important requirement in the development autonomous robot. Many localization algorithms for wheeled robots using various sensors have been proposed. In this article, we present a visual localization algorithm for a small home-use robot pet (legged robot). A low-resolution camera is equipped on the robot as the only sensor for localization. Challenges of visual localization for legged robots include: (1) Unmodeled motion errors due to leg slippages are common in legged robot, (2) as the oscillated walking motion of robot leads to fluctuated sensor data, the high degree of freedom of legged robot increases the complexity of the localization problem and (3) camera has limited field of view and image points lack of depth information. In the proposed algorithm, the localization for high-dimensional movement robot is modeled as an optimization. The objective function is then solved by a genetic algorithm. Approaches to (1) increase the efficiency of the search and (2) weaken the influence of noisy feature points to the localization results are presented. Results from simulations show that the proposed algorithm is able to localize the legged robot accurately and efficiently.

In this paper, robust human detection is investigated by fusing the stereo and infrared thermal images for effective interaction between humans and socially interactive robots. A scale-adaptive filter is first designed for the stereo vision system to detect human candidates. To eliminate the difficulty of the vision system in distinguishing human beings from human-like objects, the infrared thermal image is used to solve the ambiguity and reduce the illumination effect. Experimental results show that the fusion of these two types of images gives an improved vision system for robust human detection and identification, which is the most important and essential component of human robot interaction.