Applied Mechanics and Materials Vols. 373-375

Abstract: Coordinate axes have been used extensively in three-dimensional (3D) GIS processing to help visualize the location of points directly. Meanwhile, AxSceneControl is a visual control provided by ESRI^® ArcGIS^® Engine, which is suited to generating perspective scenes that allow user to navigate and interact with 3D feature and raster data. However, there is no direct solution given by ArcGIS Engine to set Coordinate axes in 3D scenes. In this paper, we proposed and implemented an approach to generate and customize coordinate axis in AxSceneControl. By this method, up to five coordinate axes, named as x axis, y axis, z axis, top axis and right axis, can be drawn based on the layer that is loaded to the AxSceneControl. Furthermore, users can modify the properties of coordinate axes such as changing the style of degree scale, determining which axis will be shown and resizing the degree density to achieve simple coordinate axis.

Abstract: In order to restore degraded image quality under foggy conditions, this paper proposes an improved defogging algorithm based on dark color knowledge in color space. According to the scene depth relationship between objects in a foggy image, the algorithm can repair the dark color figure, in which the grey scale line distance is used to judge edges, the most similar point of dark colors is selected as new value of dark value based on the character of the adjacent dark color, and the dark color image is used to obtain atmospheric light value automatically. Finally, the defogging purposes have been achieved based on the inverse process of the model. Experimental results show that improved algorithm effectively improves the visual effect of fog-degraded images.

Abstract: The video images of human face tracking and recognition is a hot research field of biometric recognition and artificial intelligence in recent years. This paper presents an automatic face tracking and recognition system, which can track multiple faces real-timely and recognize the identity. Aiming at Adaboost face detection algorithm is easy to false detection, presents a fusion algorithm based on Adaboost face detection algorithm and Active Shape Model. The algorithm is not only detect face real-timely but also remove the non-face areas; A multi thread CamShift tracking algorithm is proposed for many faces interlaced and face number of changes in the scene . Meanwhile, the algorithm also can identify the faces which have been tracked in the video. The experiment results show that the system is capable of improving the accurate rate of faces detection and recognition in complex backgrounds, and furthermore it also can track the real-time faces effectively.

Abstract: We address the problem of how to intuitive interacts with 3D graphic system by 2D devices, and present a general interactive framework for developing human computer interaction interface for graphical system. We consider the subclass of interaction situations in which 2D devices such as mouse and windows intuitively control scene observation, select and manipulate 3D object. Based above, an object-oriented software architecture is presented to construct the virtual world, OpenGL is used to render and control 3D graphic scene and Visual C++ 6.0 platform and Windows system are combined to build the underlying structure of the software. First, compare to OpenGL select mechanism, a novel select mechanism is presented. Second, with common 2D devices, we introduced an intuitive and precise mechanism for scene observation and object manipulation. Experimental examples prove that our interactive technology is feasible and practical.

Abstract: The algorithm of blind detection on DWT (Discrete Wavelet Transform) digital watermarking of still image is proposed to overcome the lower detection rate and higher false alarm rate problem. The algorithm utilizes KFDA (Kernel Fisher Discrimination Analysis) theory. With the help of research results of blind detection on DCT digital watermarking, the algorithm passes the test information by stochastic resonance system so as to amplify weak signals. Then the algorithm chooses suitable sample vector by computation. KFDA theory, a kind of learning machine with high precision is used to realize blind detection. Both theoretical analysis and simulation results show that the algorithm improves detection probability at low embedding strength. At the same time the algorithm also decreases false alarm rate.

Abstract: This paper presents a novel image acquisition technology for raised characters. First, laser stripes modulated with height information of raised character are captured using the 3D vision technique; then, laser stripes are combined into a grating image, in which are complete character images with grating background; Finally, a rational design Gabor kernel is applied to filter the grating image. Through this way, the background was removed and the grayscale images of the raised characters are reserved. Experiments show that the proposed method can get the well-separated character images. Also, it is more simple and efficient than the existing methods.

Abstract: In realistic image processing, it is a problem of image foreground extraction. For a large number of color image processing, an important requirement is the automation of the extraction process. In this paper, by automatically setting foreground seed, we improve the image existing segmentation algorithm; by automatically searching image segmentation region, we accomplish image segmentation with the GrabCut algorithm, which is based on Gaussian Mixture Model and boundary computing. The improved algorithm in this paper can achieve the automation of image segmentation, without user participation in the implementation process, at the same time, it improves the efficiency of image segmentation, and gets a good result of image segmentation in complex background.

Abstract: One’s head pose can be estimated using face images. The hidden manifold of head pose in the high dimensional space can be successfully embedded into a 2 dimensional space using Kernel Principal Component Analysis (KPCA). A pose curve is gotten using KPCA train samples and new pose image is projected onto this curve. The pose angle can be estimated using interpolation method. The disadvantage of traditional linear method is conquered by using 2-D KPCA and the experimental results that the method is effective to estimate head poses. The kernel functions effects on estimation accuracy are also discussed.

Abstract: To provide with an effective method for non-photorealistic rendering for computer generated images with halftoning artistic appearances from 2D images motivates our work in this paper. The methods proposed in this paper are inspired by improved error diffusion method and image enhancement, and the whole diffusion algorithm is based on the average threshold. Firstly, source image should be transferred to the gray image. Then, error diffusion, spread component and parameter confirming can be used to obtain more details. Experimental results show that the proposed method can simulate the halftoning effect in real time, and blemishes of the results can be eliminated.

Abstract: Human face detection is the first critical step of face recognition system. This paper proposed a face detection method based on skin color feature. Firstly, the method of building a skin color feature from RGB to YCbCr and extracting skin color region according the chrominance similarity was used to extract the face gray image. Secondly, image smoothness and image binarization were used to receive the binary image, then mathematical morphology operators were used to eliminate the binary images noise and disturbance. At last, human face regions are detected through projection operation. The result of experimentation affirms that the method is efficient to detect human face.