No abstract

In real security scenarios, gait data may be highly distorted due to perspective effects and there may be significant change in appearance, orientation and occlusion between different measurements. To deal with this problem, a new identification technique is proposed by reconstructing 3D models of the walking subject, which are then used to identify subject images from an arbitrary camera. 3D models in one gait cycle are aligned to match silhouettes in a 2D gait cycle by estimating the positions of a 3D and 2D gait cycles in a 3D space. This allows the gait data in a gallery and probe share the same appearance, perspective and occlusion. Generic Fourier Descriptors are used as gait features. The performance is evaluated using a new collected dataset of 17 subjects walking in a narrow walkway. A Correct Classification Rate of 98.8% is achieved. This high recognition rate has still been achieved using a modest number of features. The analysis indicate that the technique can handle truncated gait cycles of different length and is insensitive to noisy silhouettes. However, calibration errors have a negative impact upon recognition performance.

Tumors in brains are caused by the unregulated emergence of tissue cells inside the brain. The early diagnosis and determining the precise location of the tumor in magnetic resonance imaging (MRI) and its size are essential for the teams of physicians. Image segmentation is often considered a preliminary step in medical image analyses. K-means clustering has been widely adopted for brain tumor detection. The result of this technique is a list of cluster images. The challenge of this method is the difficulty of selecting the appropriate cluster section that depicts the tumor. In this work, we analyze the influence of different image clusters. Each cluster is then split into the left and right parts. After that, the texture features are depicted in each part. Furthermore, the bilateral symmetry measure is applied to estimate the cluster that contains the tumor. Finally, the connected component labeling is employed to determine the target cluster for brain tumor detection. The developed technique is applied to 30 MRI images. The encouraging accuracy of 87% is obtained.

No abstract