Abstract: An image data processing system includes an image input module, a bit map generation module and an encoding module. The image input module receives an image and secret information and obtains at least one gray-level image according to image. The bit map generation module generates a bit map image according to gray-level image and generates an ordered dithering block truncation coding (ODBTC) image according to bit map image and gray-level image. The encoding module generates an encoded gray-level image according to bit map image and secret information, wherein the encoding module divides gray-level image into image blocks, finds candidate blocks with close black and white pixel numbers among the image blocks, selects high frequency blocks belong to the high frequency region of the image from candidate blocks, and randomly embeds secret information into high frequency blocks to generate encoded gray-level image with secret information embedded therein.

Abstract: A system for hiding secret data in halftone images is disclosed. The system includes an image input module, a halftoning processing module and an image output module. The image input module reads an original image data and a secret image data. The halftoning processing module is coupled to the image input module and receives the original image data and the secret image data, performs a first error diffusion process to the original image to generate a first halftone image and performs a second error diffusion process to the original image data and the secret image data to generate a second halftone image. The image output image module is coupled to the halftoning processing module and outputs the first halftone image and the second halftone image, wherein the second error diffusion process comprises adding or subtracting a variable noise strength to or from each pixel of the secret image.

Abstract: A blind wavelet-based watermarking method is provided to extract one or more embedded watermarks form one or more high subbands of a watermarked image generated by 1-level or 2 level wavelet transform. One or more least-mean-square (LMS) filters are trained to predict the data sets in the high subbands of an original image by converting a low subband of the watermarked image. Therefore, the one or more embedded watermarks could be extracted by comparing the predicted data sets in the high subbands with data sets in corresponding subbands of the watermarked image.

Abstract: An image compression method and a device thereof are provided. In the image compression method, a block image having a plurality of pixels is received. In the block image, an adjusted value of a first pixel is compared with a threshold value, and the adjusted value of the first pixel is accordingly quantized as a first parameter value or a second parameter value so as to obtain a quantized value of the first pixel. Next, an error value between the adjusted value of the first pixel and the quantized value of the first pixel is distributed to at least one second pixel adjacent to the first pixel according to an error diffusion kernel. The first pixel and the second pixel are respectively one of the above pixels. Therefore, the quality of the compressed image is improved and the influence of the blocking effect is reduced.

Abstract: A digital halftoning method and a method for constructing a class tiling map are disclosed. The present invention utilizes to arrange class matrixes in a staggering form such that adjacent class matrixes exist a staggering shift. Also, the class tiling map is constructed in this way. Then, a dot diffusion procedure is performed to an original image with the class tiling map so as to generate a halftone image corresponding to the original image. The present invention is capable of solving the problem of a periodic appearance caused by traditional dot diffusion.

Abstract: Text processing includes: segmenting received text based on a lexicon of smallest semantic units to obtain medium-grained segmentation results; merging the medium-grained segmentation results to obtain coarse-grained segmentation results, the coarse-grained segmentation results having coarser granularity than the medium-grained segmentation results; looking up in the lexicon of smallest semantic units respective search elements that correspond to segments in the medium-grained segmentation results; and forming fine-grained segmentation results based on the respective search elements, the fine-grained segmentation results having finer granularity than the medium-grained segmentation results.

Abstract: An object detection method and an apparatus thereof are provided. In the object detection method, a plurality of images in an image sequence is sequentially received. When a current image is received, a latest background image is established by referring to the current image and the M images previous to the current image, so as to update one of N background images, wherein M and N are positive integers. Next, color models of the current image and the background images are analyzed to determine whether a pixel in the current image belongs to a foreground object. Accordingly, the accuracy in object detection is increased by instantly updating the background images.

Abstract: A mutual optimization method for class matrix and diffusion weighting used in a halftone image processing technique and a mutual optimization system thereof are provided. In the method, a mutual optimization of a plurality of diffusion weightings and a class matrix used in a dot diffusion process is performed based on a concept of simulated annealing in order to avoid converging to local solution, so as to ensure an image quality of a halftone image generated by the dot diffusion process. Besides, since the mutual optimization method is for a hexagonal class matrix, a block effect appearing in the halftone image can be significantly reduced.

Abstract: The present invention discloses a digital halftoning method. The method comprises steps of: (a1) dividing an original image into non-overlapping blocks; (a2) obtaining a Least-Mean-Square trained (LMS-trained) filter by comparing at least a training image and a halftone result corresponding to the training image (a3) optimizing a class matrix with the LMS-trained filter, which involves the diffused area and the diffused weightings; and (a4) processing the non-overlapping blocks by performing a dot diffusion procedure with the optimized class matrix and the diffused weightings to generate a halftone image corresponding to the original image. A detailed class matrix optimizing method as in the above-mentioned step (a3) is also disclosed.

Abstract: A method is performed to hide a watermark in two halftone images or in a single halftone image. The halftone image contains information of the hidden watermark, and each pixel value of the halftone image is associated with a corresponding offset value. The offset value is determined according to a corresponding pixel value of the watermark and a corresponding pixel value of another halftone image. The offset value is used to compensate a corresponding input value and a corresponding original error value of the halftone image.

Abstract: An image data processing system includes an image input module, a bit map generation module and an encoding module. The image input module receives an image and secret information and obtains at least one gray-level image according to image. The bit map generation module generates a bit map image according to gray-level image and generates an ordered dithering block truncation coding (ODBTC) image according to bit map image and gray-level image. The encoding module generates an encoded gray-level image according to bit map image and secret information, wherein the encoding module divides gray-level image into image blocks, finds candidate blocks with close black and white pixel numbers among the image blocks, selects high frequency blocks belong to the high frequency region of the image from candidate blocks, and randomly embeds secret information into high frequency blocks to generate encoded gray-level image with secret information embedded therein.

Abstract: A halftone image generation method used in a system including an image input module and a halftoning processing module for generating a halftone image is disclosed. First, an original image is received by the image input module and a dot diffusion process is performed to the original image to generate the halftone image using a first class matrix with a first size and a corresponding first diffused weighting matrix with a first diffused area size, wherein the first class matrix indicates a processing order of the dot diffusion process and the first class matrix with the first size, the first diffused area size and the corresponding first diffused weighting matrix with the first diffused area size are optimized results determined in advance by the halftoning processing module according to class matrixes of different sizes and diffused areas of different sizes.

Abstract: An ultrasonic diagnostic imaging system produces an extended field of view (EFOV) image. A 3D imaging probe is moved along the skin of a patient above the anatomy which is to be included in the EFOV image. As the probe is moved, images are acquired from a plurality of differently oriented image planes such as a sagittal plane and a transverse plane. As the probe is moved the image data of successive planes of one of the orientations is compared to estimate the motion of the probe. These motion estimates are used to position a succession of images acquired in one of the orientations accurately with respect to each other in an EFOV display format. The display format may be either a 2D EFOV image or a 3D EFOV image.

Abstract: An ultrasonic diagnostic imaging system produces an extended field of view (EFOV) image. A 3D imaging probe is moved along the skin of a patient above the anatomy which is to be included in the EFOV image. As the probe is moved, images are acquired from a plurality of differently oriented image planes such as a sagittal plane and a transverse plane. As the probe is moved the image data of successive planes of one of the orientations is compared to estimate the motion of the probe. These motion estimates are used to position a succession of images acquired in one of the orientations accurately with respect to each other in an EFOV display format. The motion estimates are also used to display a graphic on the display screen which indicates the progress of the scan to the user as the probe is being moved. The progress may be indicated in terms of probe velocity, distance traveled, or the path traversed by the moving probe.

Abstract: An image processing method combining compression and watermark techniques is provided. In the method, a block is taken as a unit, and two ordered dither arrays mutually conjugate are used to code an original image, so as to insert information of a watermark into the original image during the coding process. Accordingly, the method generates a compressed image gives consideration to the compression ratio, the image quality and the data safety. On the other hand, the method also provides a procedure to extract the watermark from the compressed image, in order to verify whether the compressed image has been modified illegally.

Abstract: A system for hiding secret data in halftone images is disclosed. The system includes an image input module, a halftoning processing module and an image output module. The image input module reads an original image data and a secret image data. The halftoning processing module is coupled to the image input module and receives the original image data and the secret image data, performs a first error diffusion process to the original image to generate a first halftone image and performs a second error diffusion process to the original image data and the secret image data to generate a second halftone image. The image output image module is coupled to the halftoning processing module and outputs the first halftone image and the second halftone image, wherein the second error diffusion process comprises adding or subtracting a variable noise strength to or from each pixel of the secret image.

Abstract: An image compression method is provided. First, an image including a plurality of blocks is received and each block includes a plurality of pixels. Next, an order dithering array within a preset range is generated, which includes a plurality of elements. Then, values of the elements in the order dithering array are respectively mapped to a plurality of mapping values according to a gray scale range of each block. In each block, each pixel value is quantized to a first digital value or a second digital value according to a result of comparing each pixel value with the corresponding mapping value. Hence, a bit rate of the image is efficiently compressed, and features of the original image are preserved inside the compressed image.

Abstract: A method is performed to hide a watermark in two halftone images or in a single halftone image. The halftone image contains information of the hidden watermark, and each pixel value of the halftone image is associated with a corresponding offset value. The offset value is determined according to a corresponding pixel value of the watermark and a corresponding pixel value of another halftone image. The offset value is used to compensate a corresponding input value and a corresponding original error value of the halftone image.

Abstract: An image compression method and a device thereof are provided. In the image compression method, a block image having a plurality of pixels is received. In the block image, an adjusted value of a first pixel is compared with a threshold value, and the adjusted value of the first pixel is accordingly quantized as a first parameter value or a second parameter value so as to obtain a quantized value of the first pixel. Next, an error value between the adjusted value of the first pixel and the quantized value of the first pixel is distributed to at least one second pixel adjacent to the first pixel according to an error diffusion kernel. The first pixel and the second pixel are respectively one of the above pixels. Therefore, the quality of the compressed image is improved and the influence of the blocking effect is reduced.

Abstract: A blind wavelet-based watermarking method is provided to extract one or more embedded watermarks form one or more high subbands of a watermarked image generated by 1-level or 2 level wavelet transform. One or more least-mean-square (LMS) filters are trained to predict the data sets in the high subbands of an original image by converting a low subband of the watermarked image. Therefore, the one or more embedded watermarks could be extracted by comparing the predicted data sets in the high subbands with data sets in corresponding subbands of the watermarked image.