Jian-Wei Zhang

Image registration based on generalized and mean Hausdorff distances

A new method to register two un-identical imaging representation of the same objects based on generalized and mean Hausdorff distances is proposed. Firstly two image contours are registered by minimizing mean Hausdorff distance used as cost function, through two-dimensional translation and rotation with simulate anneal algorithm. The registration is often inaccurate due to the location-independent difference of two contours. Generalized Hausdorff distance was analyzed to ascertain the excess of the floating image over the model image. Then the new floating image subtracted the excess points is registered to the model image. Accurate registration was attained after several iterations.

Computational spatiotemporal analysis identifies WAVE2 and cofilin as joint regulators of costimulation-mediated T cell actin dynamics

Live-cell microscopy determines the relative contributions of different signaling pathways to actin reorganization in T cells. Imaging T cell actin dynamics T cells must receive signals through the T cell receptor (TCR) and the costimulatory receptor CD28 to become fully activated. Critical to this process is the reorganization of plasma membrane actin at the immunological synapse, the interface between a T cell and an antigen-presenting cell. Roybal et al. imaged actin and fluorescently tagged actin regulatory proteins in T cells activated through the TCR in the absence or presence of CD28 signaling. Computational image processing to normalize differences in cell shape enabled tracking of the fluorescent proteins. The regulatory proteins WAVE2 and cofilin were efficiently recruited to the immunological synapse only when both TCR and CD28 signaled. Constitutive activation of either protein in TCR-stimulated T cells enabled normal actin reorganization even when CD28 signaling was blocked. This combination of imaging and computational analysis could be applied to other systems to determine the spatiotemporal dynamics of signaling molecules. Fluorescence microscopy is one of the most important tools in cell biology research because it provides spatial and temporal information to investigate regulatory systems inside cells. This technique can generate data in the form of signal intensities at thousands of positions resolved inside individual live cells. However, given extensive cell-to-cell variation, these data cannot be readily assembled into three- or four-dimensional maps of protein concentration that can be compared across different cells and conditions. We have developed a method to enable comparison of imaging data from many cells and applied it to investigate actin dynamics in T cell activation. Antigen recognition in T cells by the T cell receptor (TCR) is amplified by engagement of the costimulatory receptor CD28. We imaged actin and eight core actin regulators to generate over a thousand movies of T cells under conditions in which CD28 was either engaged or blocked in the context of a strong TCR signal. Our computational analysis showed that the primary effect of costimulation blockade was to decrease recruitment of the activator of actin nucleation WAVE2 (Wiskott-Aldrich syndrome protein family verprolin-homologous protein 2) and the actin-severing protein cofilin to F-actin. Reconstitution of WAVE2 and cofilin activity restored the defect in actin signaling dynamics caused by costimulation blockade. Thus, we have developed and validated an approach to quantify protein distributions in time and space for the analysis of complex regulatory systems.

Image Authentication Resilient to Translation, Rotation and Scaling

This paper presents an image content authentication algorithm resilient to geometric and some general operations. The original image is normalized into a standard prototype form according to the blind normalization arithmetic, and then the smooth-component and the edge-character are drawn from the prototype image with the dyadic wavelet transform multi-scale edge detection. After coded with 64 bits, the smooth-component along with the serialized edge-character is embedded in the original image as watermark. When verifying an image, the same feature is computed and the results are compared with the watermark extracted from the test image to confirm whether the content has been tampered. The experiments prove that this authentication algorithm can effectively detect the vicious tamper. Moreover it can tolerate not only the distortion produced by compression, filter and some other image processing but also the geometrical transformations

Segmentation of overlapping cells in cervical smears based on spatial relationship and Overlapping Translucency Light Transmission Model

Overlapping cell segmentation in cervical smear images is a difficult task due to the shape multiformity and color proximity of the cells. In this paper, we propose a segmentation approach by using the spatial relationship of the non-overlapping and overlapping areas as well as Overlapping Translucency Light Transmission Model (OTLTM) to segment the overlapping cells in these images. The spatial relationship, which denotes the overlapping area locates in the middle ground of the non-overlapping areas, reflects the overlapping area can be accurately gained by the precise non-overlapping areas. After removing the background by threshold technique, a fragmentation method by using mean shift and watershed is adopted to divide the overlapping cells into fragments according to the similarity of their colors. The fragments belong to a single-tier individual cell, or to the overlapping area between two cells. We firstly construct the initial fragment collections of non-overlapping areas based on the Voronoi diagram, then the initial collections are optimized by using the initial cell overlapping matrix based on the spatial relationship, and OTLTM based on Beer-Lambert law, which states the relationship between the transmittance, attenuation coefficient of a kind of material and the distance the light travels through it. The cell overlapping matrix is accurately reconstructed by the optimized set of the non-overlapping areas. We obtained the segmentation result by combining the cell overlapping matrix and the optimized set of the non-overlapping areas. The experimental results show that the proposed method can give an impressive performance. Besides cervical smear images, these proposed techniques can be utilized in segmenting translucent objects from other kinds of images. We obtain the fragments by coarse segmentation and base these fragments to achieve the result.We propose Overlapping Translucency Light Transmission Model.We derive overlap color measurement formula, and provide overlapping fragment judges theory.We propose spatial relationship of the non-overlapping and overlapping areas.Our method can also be applied in segmentation of translucent object from other images.

A new segment method for segment-based image coding

A new segment algorithm based on the pixels intensity and spatial information is proposed in this paper. The algorithm has three main steps. The first step is image preprocessing. After the points in the neighborhood of each image pixel are classified, their distributing situation are used to estimate the image pixels location, in the center or near the boundary of a region. Each pixels gradient magnitude is adaptively adjusted according to its location. The second step is image segmentation. Watershed transform is performed on the adjusted gradient magnitude image to obtain the initial segmented regions. The last step is regions merging. Pairs of adjacent regions are merged if they have small cost function. For each pair of adjacent regions, the cost function is calculated according to the strength of their shared boundaries, the ratio of smaller region area to shared boundary length, and the homogeneity. Experimental results indicate that this algorithm provides a good segmentation for segment-based image coding.

Adaptive segmentation of cervical smear image based on GVF Snake model

The key of a computer-assisted diagnosis system for screening of cervical cancer is the accurate segmentation of cells. In this paper, an adaptive segmentation algorithm based on GVF Snake model is proposed to separate the nucleus from cervical smear model. We set the parameters of the model, and then use the model to segment the cervical cells based on the initial contour of nuclei. The segmentation results are evaluated, if the results meet the criterion, the segmentation process finishes, otherwise we adjust the parameter which is a weight in energy function when calculating GVF field and perform the segmentation procedure again. Adjusting the parameter by evaluating the segmentation results is an adaptive and sensible method with regard to the automatic segmentation. The experiment results show the effectiveness of the proposed approach in images having inconsistent staining and poor contrast.

Improved codebook model based on spatio-temporal context

Spatio-temporal context refers to the information of each pixels historical status and its adjacent pixels. The proposed algorithm applies the spatio-temporal context to the detection of foregrounds. It is based on the codebook model. For each pixel, a weight value is calculated according to the spatio-temporal context of this pixel to influence the detecting conditions. The algorithm can make the detecting results more accurate, especially in the interference regions such as the waving trees and the sudden illumination.

Improved CamShift tracking algorithm based on motion detection

The traditional Continuously Adaptive Mean Shift Algorithm (CamShift) is widely used, but its drawback is unsatisfactory performance due to counting aU pixels when calculating the color histogram and back projection using a rectangular box to select the target. We propose an improved CamShift Algorithm based on motion detection. When calculating the color histogram of a target within the rectangular box, it adds a mask layer to remove background pixels around targets. When calculating the back projection, it adds a mask layer to remove all background pixels within the window to eliminate the interference of similar color in the background. The experimental results show that this improved algorithm utilizes the color feature better and keeps the tracking right even when background interference exists.

A background modeling method for videos based on weighted statistical classification

In the field of intelligent video surveillance, foreground detection, moving target tracking and target recognition are the key technologies. They play an important role in target behavior analysis and understanding. In this paper a background modeling method based on weighted statistical classification is proposed. As a non-parametric background model, it uses several state categories to express multiple states of a background pixel. It does not require the background pixels to obey Gaussian distribution and needs no training. The weights are updated according to the matching history of the background pixel. The background state is determined by a threshold. Experiment results show that it obtains excellent detection results and real-time detection speed in complex scenes.

Detection of abnormal nuclei in cervical smear images based on visual attention model

A novel idea for detecting abnormal nuclei in cervical smear images is presented. The suspect cells often appear with different externals from surrounding normal ones. Therefore, we are able to find them and focus on processing them instead of segmenting all nuclei in the images. This method combines the bottom-up attention mechanism and the top-down target-driven detection method. We extract both direction and brightness features of the image and construct a saliency map which is linearly combined with the high response area detected using annular template matching method. Then, we use an inhibition-of-return as well as winner-take-all mechanism to detect the regions of interest one by one. This process will be followed by segmentation and recognition of the found nuclei. The satisfactory results on extraction and computing speed show that this model can extract the abnormal nucleus regions without processing other part of the image.