Search

Fundus images have been widely used in the diagnosis of retinopathy and cardiovascular diseases. However, because of the movement of patients' eyes and limitation of medical equipments, the quality of fundus images may be low sometimes. In this paper, we propose vessel enhancement method for a low-contrast and blurred image based on multi-scale morphological top-hat transformation, and the combination of Gabor and matched filter. The underlying rationale of this proposed method is to make use different kinds of information to improve blood vessels on retinal fundus images with poor quality. Experiments show that vessels enhanced by our proposed method are much clearer than the ones using original multi-scale morphological top-hat transformation, Gabor filter or matched filter. The results confirm that the quality of vessel enhancement can be improved by combining these methods.

Depth estimation from a sequence of images is a challenging problem in computer vision research. One of the well-known solutions is the depth from focus. However, the drawbacks of this method are the tradeoff between spatial resolution and robustness, and failure in textureless regions. In this paper, a novel approach of depth from focus with multiple images is proposed to improve the two shortcomings. By employing the mean shift segmentation before the step of building Markov random field, the result of segmentation serves as adaptive window for DFF. The edges of the recovered depth map are guaranteed to align with the edges of the original image. After the initial estimation of depth, the hierarchical Markov random field is generated to expand the area to extract depth information according to the structure of the scene. In this way, the experiments show that depth can extract from the textureless regions to some extent.

RFID has been widely adopted in access control as an identity identification technology. This system can be combined with the face recognition technique to avoid the RFID is used by unauthorized people. However, this approach also suffers a drawback that someone may try to deceive the security system by presenting a photo of the real card owner. In this paper, we propose a method ensuring that such kind of spoof can be rejected. After a frontal face has been detected, the facial object will be tracked by optical flow algorithm until a profile is detected. Experimental results show that the proposed mechanism can improve the validity of the access control system effectively.

Radio frequency identification (RFID) technology has been widely adopted in access control system. However, the people holding the RFID card passing through the access control may not be the authorized one. Therefore, an access control system combining RFID technology and face recognition based on neural network is presented in this work. The system recognizes the face of the person holding the RFID card and denies access if they do not match. We adopt a Radial Basis Function Neural Network (RBFNN) to learn the face of authorized card holders and save the parameters of RBFNN only. This could reduce storage when the number of card holders getting large. Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) features are extracted to reduce the dimensions of face image data. The Localized Generalization Error Model (L-GEM) is adopted to train a RBFNN for better generalization capability. The face recognition system is first evaluated by benchmarking ORL face image database. The whole access control system is then tested in a real environment. Experimental results show that the proposed method has a good performance and could improve the security of RFID access control.

RFID identification has been widely adopted in access control. This kind of card or tag based approaches has a major drawback that anyone could get access with the card. In this work, we propose a neural network based face recognition system as the second access control to make sure the person granted access matches the ID on the RFID card. In this preliminary work, the face of accessing person is detected in video stream and we extract the Scale Invariant Feature Transform (SIFT) features from a face image. To enhance the generalization capability of the face recognition, we introduced the Localized Generalization Error Model (L-GEM) to train the Radial Basis Function Neural Network (RBFNN) for face recognition. Experimental results show that the proposed method could identify person that matches the RFID access card or not in a high probability.