Abstract: In a defect detection apparatus, images of first to third inspection areas on a substrate are picked up to acquire first to third images. A positional difference acquisition part (51) acquires a first difference vector between the first image and the second image and a second difference vector between the second image and the third image. A differential image generation part (52) generates a first differential image between the first image and the second image while adjusting a position of the first image to the second image on the basis of the first difference vector and a second differential image between the second image and the third image while adjusting a position of the second image to the third image on the basis of the second difference vector.

Abstract: A method for processing wafers includes learning a first pattern at a de-skew site on a first wafer layer, saving the first patterns in a recipe for de-skewing wafers, learning a second pattern at the de-skew site a second wafer layer, and saving the second pattern in the same recipe for de-skewing wafers. Learning the first pattern may include determining a score of uniqueness for the first pattern. The method further includes finding the de-skew site on the second wafer layer using the first pattern before learning the second pattern. Finding the de-skew site includes determining a score of similarity between the first pattern and the second pattern. Learning the second pattern is performed when the score of similarity is less than a threshold value. A recipe for de-skewing wafers includes multiple patterns of a de-skew site of a wafer, wherein the patterns include a first pattern at the de-skew site on a first wafer layer and a second pattern at the de-skew site on a second wafer layer.

Abstract: A method of JPEG compression of an image frame divided up into a plurality of non-overlapping, tiled 8×8 pixel blocks Xi. A global quantization matrix Q is determined by either selecting a standard JPEG quantization table or selecting a quantization table such that the magnitude of each quantization matrix coefficient, Q[m,n] is inversely proportional to the aggregate visual importance in the image of the corresponding DCT basis vector. Next a linear scaling factor Si is selected for each block, bounded by user selected values Smin and Smax. Transform coefficients, Yi, obtained from a digital cosine transform of Xi, are quantized with global table Smin Q while emulated the effects of quantization with local table Si Q and the quantized coefficients Ti[m,n] and global quantization table Smin Q are entropy encoded , where Smin is a user selected minimum scaling factor, to create a JPEG Part 1 image file.

Abstract: A traffic lane marking line recognition system for vehicle including a traffic lane marking line recognizing device configured to recognize at least a traffic lane marking line in broken (e.g., white) line on a road surface in a camera photographed image, and an image compositing device which composites images photographed at different time points to elongate the traffic lane marking line in the photographed image, wherein the images are composited in a processing stage in such a manner that no change will be imparted to at least a shape of the traffic lane marking line in the photographed image in a traffic lane marking line recognition processing, specifically in a processing stage prior to edge detection. With this, a distant traffic lane marking line is prevented from being chipped away when compositing the photographed images, thereby enabling accurate and unerring recognition of the traffic lane marking line.

Abstract: In the traffic lane marking line recognition system for vehicle having an image compositing device for compositing images photographed at different time points by a camera to elongate a traffic lane marking line in the photographed image, a fore-aft direction angle of the vehicle relative to the traffic lane marking line and a distance of the vehicle from the line in each of the images to be composited are detected to determine deviations of the angles and distances between the images; and the images are composited after having been corrected such that the angles and distances of the vehicle relative to the line are equal to each other based on the determined deviations of the angles and distances. With this, the traffic lane marking lines are elongated in appearance in the composited image due to the correction, thereby enabling accurate and unerring recognition of the traffic lane marking line.

Abstract: The present invention provides a system, method and computer program product for improving quality of images. The pixels of the input image are used to generate two lists of pixels based on two different scale space values. Thereafter, the two pixels lists are processed to obtain an output image with improved quality.

Abstract: The present invention relates to a method and system for foreground segmentation in which frames of a video sequence are analyzed in the transform domain to determine one or more features. The features are used to model the background. The background can be modeled as a single Gaussian model with a mean and variance of the features. A current frame is segmented by determining if one or more features of the current frame analyzed in the foreground domain satisfy a threshold between the background model. The threshold value can be based on the mean and/or variance of features. During the segmentation, the mean and variance can be updated based on previous corresponding values and current features to adaptively update the background model. In one embodiment, the frames are divided into a plurality of blocks. A transform is used to analyze the blocks in the transform domain. For example, the transform can be a discrete cosine transform (DCT).

Abstract: A traffic lane marking line recognition system for vehicle includes an image compositing device which composites images photographed at different time points by a camera to elongate a traffic lane marking line such as a white line in the photographed image based on a detected travel speed of the vehicle. At least one image, among the plurality of images, photographed at a past time point earlier by a time interval determined in accordance with the vehicle speed from a reference time point, when defining a photographed time point of a given one of the images to be composited as the reference time point, is selected as the image to be composited with the given one of the images. By selecting the image to be composited in accordance with the vehicle speed, the traffic lane marking lines can be recognized unerringly.

Abstract: A traffic lane marking line recognition system for a vehicle includes an image compositing device which composites images photographed at different time points by a camera to elongate a traffic lane marking line such as a white line in the photographed image, the brightness of the photographed images is detected, the brightness is corrected, taking one image among the plurality of images as a reference, such that the brightness of the plurality of images is equal to each other, and the plurality of images are composited after the brightness has been corrected. By making the brightness of the images to be composited equal to each other, it becomes possible to recognize the traffic lane marking line accurately and unerringly.

Abstract: A low-frequency, temporally asymmetric watermark carrier decreases in luminance at a slower rate than it increases in luminance. This allows the carrier to be embedded, e.g., in a video work, in a manner that renders the resultant watermark virtually invisible to the human eye. Moreover, the resultant watermark is substantially resistant to common pirating techniques in which high-frequency visual content (e.g., including watermark data) tends to be filtered out by piracy processing.

Abstract: A red-eye detection device includes a red-eye candidate detection section, a red-eye pair specification section, and an isolated red-eye candidate check section. The red-eye candidate detection section detects a red-eye candidate that can be estimated to be a red-eye, by identifying a feature of the pupil that has a region displayed red from among features of an image. The red-eye pair specification section specifies the red-eye candidate as a pair of a right red-eye candidate equivalent to a right eye and a left red-eye candidate equivalent to a left eye, using information on a face region obtained from the image. The isolated red-eye candidate check section checks whether an isolated red-eye candidate, detected by the red-eye candidate detection section but not specified as the pair by the red-eye pair specification section, is a true red-eye.

Abstract: Disclosed are a method for determining a dark condition in an image sensor to prevent a color noise and a method for interpolating colors. The present invention forms a window with a size of 5×5 array and determines a local color distribution based on a center pixel by using the window with the size of 5×5 array, thereby determining a dark condition with respect to the local color distribution. Thus, it is possible not to degrade chroma with respect to a whole image but locally degrade chroma, thereby producing a black and white image. Through differently and locally performing a color interpolation according to a dark condition determination, it is possible to correct an error of an imperfect black possibly generated during an expression of an imperfect black and white image due to the color interpolation and to reduce a color noise in the corresponding portion.

Abstract: An implementation of a technology is described herein for recognizing the perceptual similarity of the content of digital goods. At least one implementation, described herein, introduces a new hashing technique. More particularly, this hashing technique produces hash values for digital goods that are proximally near each other, when the digital goods contain perceptually similar content. In other words, if the content of digital goods are perceptually similar, then their hash values are, likewise, similar. The hash values are proximally near each other. This is unlike conventional hashing techniques where the hash values of goods with perceptually similar content are far apart with high probability in some distance sense (e.g., Hamming). This abstract itself is not intended to limit the scope of this patent. The scope of the present invention is pointed out in the appending claims.

Abstract: An image capturing apparatus provided to a movable body for capturing a clear image without blurring even if it is the case where it is moving.

Abstract: Edges in a block of a digital image are detected by determining an entropy of a histogram of the block.

Abstract: Methods and systems for compression of digital images (still or motion sequences) are provided wherein predetermined criteria may be used to identify a plurality of areas of interest in the image, and each area of interest is encoded with a corresponding quality level (Q-factor). In particular, the predetermined criteria may be derived from measurements of where a viewing audience is focusing their gaze (area of interest). In addition, the predetermined criteria may be used to create areas of interest in an image in order to focus an observer's attention to that area. Portions of the image outside of the areas of interest are encoded at a lower quality factor and bit rate. The result is higher compression ratios without adversely affecting a viewer's perception of the overall quality of the image.

Abstract: One embodiment disclosed relates to a method of detecting defects in a microminiature pattern repeated over a particular surface of an object. The method includes providing an image detector capable of resolving intensity data for each pixel of an array of m×n pixels of an image projected thereupon. An area of the surface of the object is illuminated with incident electrons, radiation from the illuminated area is detected, and image data from the detected radiation is retrieved into a data processor. If necessary, the image data is rotated for alignment purposes. The scale of the image data is then adjusted such that a dimension of a cell of the repeated pattern in the image data is equivalent to a distance across an integer number of pixels of the m×n pixel array. With the scale adjusted, then features in one cell of the repeated pattern are compared with features in another cell of the repeated pattern to detect said defects.

Abstract: A method and system is configured to characterize regions of an environment by the likelihoods of transition of a target from each region to another. The likelihoods of transition between regions is preferably used in combination with conventional object-tracking algorithms to determine the likelihood that a newly-appearing object in a scene corresponds to a recently-disappeared target. The likelihoods of transition may be predefined based on the particular environment, or may be determined based on prior appearances and disappearances in the environment, or a combination of both. The likelihoods of transition may also vary as a function of the time of day, day of the week, and other factors that may affect the likelihoods of transitions between regions in the particular surveillance environment.

Abstract: A method and an related apparatus for performing a white balance operation on an image with pixels, wherein global colour gain values common for substantially all pixels of the image are first determined. Then local colour gain values for individual pixels of the image are determined based on analyses of a local window comprising said individual pixel and a number of adjacent pixels. Thereafter, final colour gain values for individual pixels of the image are determined as a weighted average of the global and the local colour gain values, and a white balance operation is performed on the image using the final colour gain values.

Abstract: The present invention permits to determine an image characteristic amount for efficiently searching an image without affecting a bad influence upon searching accuracy. First image characteristic amount data is extracted. Second image characteristic amount data is extracted. Then, a distance between two image characteristic amount data is calculated. Further, the calculated distance is compared with a threshold value. If the distance is smaller than the threshold value, the image characteristic amount data having smaller number of coefficients which is advantageous because of smaller data amount is selected, and, if otherwise, the image characteristic amount data having greater number of coefficients is selected.