Applied Mechanics and Materials Vols. 373-375

Abstract: The auditory filter-bank is the key component of auditory model, and its implementation involves a lot of computations. The time spent by an auditory filter-bank to finish its work has a significant effect on the real-time implementation of auditory model-based audio signal processing systems. In this paper, a multi-rate implementation of auditory filter bank is presented. Through using low sampling rate for the filters with low centre frequency, and using high sampling rate for the filters with high centre frequency, we can greatly reduce the computation requirement.

Abstract: In this paper an improved image segmentation algorithm based on watershed transform is presented. Firstly the normalized cut method and watershed transform are explained and analyzed. Secondly the idea of the improved algorithm and the main formula are explained. We consider the area and perimeter when we merge adjacent regions. We define a new weight value and discuss the value of the parameter αand β. Finally the experiment result is presented. The new algorithm reduces the nodes and the computational demand of the common normalized cut technique.

Abstract: Due to the limited pixel numbers in single TDI CCD chips, some TDI CCD chips are combined in the focal plane to reach a sufficient swath width. Basic imagery is delivered as one image data file per strip and how to convert those sub-images to a larger, accuracy equivalent image is an essential problem in satellite images’ processing. Some key factors that affect the imaging quality of TDI CCD images are discussed firstly. It shows that discrepancies in the along-track direction which caused by different viewing angle and topographic relief are the main problem to improve the stitching accuracy. Three basic image-oriented stitching strategies are discussed. Experimental results show that the RMS of shift transformation, single affine transformation and piecewise affine transformation is 1.4 pixels, 1.3 pixels and 0.2 pixels. Piecewise affine transformation fully meets the requirements of further processing.

Abstract: This paper analyzes the structure of character by theory and experiment, and developes a hand-written character recognition system, this system recognize character by giving feature a code. It is proved feature is simple and definite ,and this system is valid.

Abstract: In a monocular video scene, in order to improve the efficiency of object tracking and counting under occlusion conditions. The article presents a scheme to automatically track and count people in a surveillance system. First, a modified Gaussian mixture model was employed to determine pedestrian objects from a static scene. To identify foreground objects by positions and sizes of foreground regions which were obtained. Moreover, the performance to track objects was improved by using the modified overlap tracker, the modified overlap tracker was used to analyze the centroid distance between neighboring objects and help object tracking and people counting in occlusion states of merging and splitting. On the experiments of tracking and counting people in three video sequences, the results show that the proposed method can improve the averaged detection ratio about 10% as compared to the conventional work.

Abstract: Due to the impact of the lens optical system and the different manufacturing processes, there could be causing the vignetting when using a CCD camera to take pictures. This paper analyzes the causes of the lens vignetting and put forward a method for reducing this error to improve the accuracy of the intensity of illumination detection. Firstly, according to the analysis of the existing research on the camera lens, concluded that there are two factors that cause vignetting. Secondly, based on the conclusion, do large amounts of test in each case. Analyze for the influence of illumination test in each case. Finally, the article put forward a set of vignetting correction method to correct the image be collected in a fixed aperture. The result shows that the method can effectively reduce vignetting measuring error and has a good stability.

Abstract: Objective To propose a PET/CT multi-planar fusion system based on PACS. Methods Firstly, the system got PET and CT slices and relevant DICOM header information from PACS workstation, preprocessed them and formed CT volume and PET volume, and then displayed the axial, coronal and sagittal planes of the two volumes by MPR. Afterwards, the results of MPR were enhanced by pseudo color processing, then were fused with the MRP results of CT correspondingly. Finally, the SUVs of PET images and the CT value of CT images were calculated according to the parameters obtained from the DICOM header, and were displayed in the lower right corner of the current sub window. Results This system can be integrated into PACS workstation expediently, can reconstruct each volume, either MPR or MIP, stably and effectively, and can fuse PET images with CT images in different planes. The resource consumption of the system is reasonable and the SUVs calculated were accurate and reliable. Conclusion The system is easy to use and has stable performance, which is helpful in the clinical diagnosis of malignant lesions as an auxiliary tool.

Abstract: The image fusion technique is applied to the field of Terahertz (THz) image processing to solve the problem that THz cameras offered low-resolution images , which do not match the human visual habits. A fusion algorithm is proposed based on region segmentation and Nonsubsampled Contourlet (NSCT). The THz image is divided into target area and background area ,which are mapped into the visible images in next step, and the different rules are designed in NSCT domain. Experimental results show that the proposed algorithm displays the target, retains the background and detail information effectively. This algorithm is more feasible than traditional fusion algorithm based on pixel and regional energy in the human eye perceives, and can be used effectively for machine vision and terahertz security area.

Abstract: In both industrial field and office building, the accurate statistic of people who enter or leave the elevator has important practical meaning in security and analysis of passenger flow. We present a binocular vision system to count the people pass by. The camera is set in the height of 2.45 meters to monitor the people overhead in order to reduce the overlap of pedestrians. The object segment and tracking method proposed in this paper show good result with the disparity map gained by the dual-camera. Dynamic promotion of threshold is used in the object segmentation. Feature matching is used to track the moving objects. The system can get the number of people accurate and timely. Experiment results show that our system has good performance under relatively complex circumstance.

Abstract: The image retrieval based on emotional keywords is required by users. But now it is short of ways to mark emotional semantic for images, especially the automatic methods, narrow and inaccurate. In this paper, we use the web images which are marked relatively comprehensively and accurately as training samples to generate emotional semantic.