Abstract: The present invention suppresses noise that is possibly generated due to threshold processing during frequency conversion. The image processing device 100 includes: a threshold processing unit 110 for executing threshold processing, by using a first parameter for conversion coefficients of respective frequency components that are obtained by frequency conversion of a block consisting of multiple pixels; a calculation unit 120 for calculating feature quantities of the constituent multiple pixels of the block; a determination unit 130 for determining a second parameter on the basis of the calculated feature quantities; and a correction unit 140 for correcting, by using the determined second parameter, the threshold-processed conversion coefficients of the respective frequency components.

Abstract: The purpose of the present invention is to reduce required calculation resources for image smoothing based on weighted least squares. The image processing device 100 includes: a first smoothing unit 110 for applying one-dimensional first smoothing processing based on weighted least squares to an input image in a first direction per a predetermined number of lines; and a second smoothing unit 120 for applying one-dimensional second smoothing processing based on weighted least squares to the first-smoothing-processed input image in a second direction, which is different from the first direction, per a predetermined block.

Abstract: In a video decoding device, a quantization step size decoding unit calculates a quantization step size that controls a granularity of the inverse quantization by, based on an image prediction, selectively using a mean value of at least a quantization step size assigned to a leftwardly adjacent neighboring image block already decoded and a quantization step size assigned to a upwardly adjacent neighboring image block already decoded or a quantization step size assigned to an image block decoded immediately before.

Abstract: A video encoding device includes: transform means 11 for transforming an image block; entropy encoding means 12 for entropy-encoding transformed data of the image block transformed by the transform means 11; PCM encoding means 13 for PCM-encoding the image block; multiplex data selection means 14 for selecting output data of the entropy encoding means 12 or the PCM encoding means 13, in a block of a block size set from the outside; and multiplexing means 15 for embedding a PCM header in a bitstream, in the block of the set from the outside block size.

Abstract: A radar image processing device includes: an estimation unit 1 for setting, as a target pixel, each pixel of a two-dimensional map image, and estimating to which type each target pixel belongs among a pixel caused by the main lobe, a pixel due to the sidelobe, and any other pixel; a pixel value replacement unit 2 for replacing the pixel value of the pixel caused by the main lobe and the pixel value of the pixel due to the sidelobe with pixel values generated in pixel value interpolation processing based on the type of each pixel estimated by the estimation unit 1 to generate a first corrected image; a speckle noise suppression unit 3 for applying speckle noise suppression processing to the first corrected image to generate a second corrected image; and an output image generation unit 4 for generating an output image, in which speckle noise and sidelobes are suppressed, by using the two-dimensional map image, the second corrected image, and the type of each pixel.

Abstract: Provided is a pattern generation device for generating a sampling pattern that could make it possible to improve the efficiency of image processing, including the processing of sampled blocks. A pattern generation device as in one embodiment of the present invention is provided with: a main generating means for generating a sampling pattern which is a pattern of positions of sampling points in an image, the sampling points indicating blocks to be used in image processing of the respective blocks, and the sampling pattern generated so that the numbers of sampling points in a plurality of rows in the sampling pattern are each equal to a reference value; and an output control means for performing a control for outputting a main sampling pattern, which is the sampling pattern generated by the main generating means.

Abstract: The present invention provides an image processing method, an image processing device, and an image processing program which require low computing costs and with which it is possible to minimize noise amplification and halo generation caused by HDR. An image processing device according to one embodiment of the present invention has: a multi-resolution image generation means for generating a multi-resolution image; a correction amount calculation means for calculating a brightness correction amount on the basis of a lowest-resolution image of the multi-resolution image, a differential image between adjacent resolutions of the multi-resolution image, and edge information calculated at each resolution of the multi-resolution image; and a noise suppression means for calculating, on the basis of the lowest-resolution image, the differential image between adjacent resolutions, the edge information, and the brightness correction amount, an image after brightness correction in which a noise component is suppressed.

Abstract: An image processing apparatus 10 includes a first generation unit 11 which generates a plurality of lower-resolution images having different row-direction resolutions by changing a resolution in a row direction of an image to be processed to a plurality of lower resolutions, a second generation unit 12 which generates a differential image by taking a difference between two of the plurality of lower-resolution images, and a computation unit 13 which computes a correction amount for the pixel values of a predetermined column of the image to be processed by use of a statistical quantity of the pixel values of a predetermined column of the differential image.

Abstract: In a video decoding device, a quantization step size decoding unit calculates a quantization step size that controls a granularity of the inverse quantization by, based on an image prediction, selectively using a mean value of at least a quantization step size assigned to a leftwardly adjacent neighboring image block already decoded and a quantization step size assigned to a upwardly adjacent neighboring image block already decoded or a quantization step size assigned to an image block decoded immediately before.

Abstract: There is provided a video encoding device capable of changing a quantization step size frequently while reducing an increase in code rate to enable high-quality moving image encoding, comprising a quantization step size encoding unit 10 for encoding a quantization step size that controls the granularity of quantization, the quantization step size encoding unit 10 includes a step size prediction unit 11 for predicting the quantization step size that controls the granularity of the quantization by using a quantization step size assigned to a neighboring image block already encoded.

Abstract: When performing noise removal of an image signal by applying the background art, there is a case where a filter coefficient that causes an edge constituting the image to become blurred against expectation due to the influence of noise in the image signal is set. The methods to solve this problem include a method for calculating a wide range of image feature values needed for correction of the filter coefficient, but this method has the problem that it requires a significant amount of calculation, thus increasing the calculation cost. To solve the above problem, the present invention is provided with an image signal input means, a wavelet transformation means, a first structure/texture separating means, a texture component degenerating means, a first combining means, an inverse wavelet transformation means, and a second structure/texture separating means. The present invention is further provided with a second texture component degenerating means, a second combining means, and an image signal output means.

Abstract: A video encoding device includes: transform means 11 for transforming an image block; entropy encoding means 12 for entropy-encoding transformed data of the image block transformed by the transform means 11; PCM encoding means 13 for PCM-encoding the image block; multiplex data selection means 14 for selecting output data of the entropy encoding means 12 or the PCM encoding means 13, in a block of a block size set from the outside; and multiplexing means 15 for embedding a PCM header in a bitstream, in the block of the set from the outside block size.

Abstract: In order to properly detect an outbreak of smoke using a photographic image, this detection device is equipped with an extraction unit, a calculation unit, and a detection unit. By using an image analysis result of an input image, that is, a photographic image captured of a scene under surveillance, the extraction unit generates an image for which a noise component is removed from the input image. By using brightness information contained in the generated image, the calculation unit calculates an attenuation factor of reflected light from a captured object in the input image. The detection unit determines whether an outbreak of smoke is present in the scene under surveillance based on the attenuation factor.

Abstract: A video encoding device includes: pixel bit length increasing means for increasing a pixel bit length of an input image based on pixel bit length increase information; transform means for transforming output data of the pixel bit length increasing means; entropy encoding means for entropy-encoding output data of the transform means; non-compression encoding means for non-compression-encoding input data; multiplexed data selection means for selecting output data of the entropy encoding means or output data of the non-compression encoding means; and multiplexing means for multiplexing the pixel bit length increase information in a bitstream, wherein a pixel bit length of an image corresponding to the output data of the entropy encoding means and a pixel bit length of an image corresponding to the output data of the non-compression encoding means are different from each other.

Abstract: A video encoding device includes a planar prediction unit using planar prediction, and a noise injecting unit for injecting pseudo-random noise into a prediction image of a planar prediction block when a reconstructed prediction error value for planar prediction is less than a threshold determined by the size of the planar prediction block. The video encoding device then uses the pseudo-random noise in linear interpolation for calculating the prediction image, based on the conditions under which a gradient distortion occurs, to suppress the gradient distortion.

Abstract: A video encoding device includes: pixel bit length increasing means for increasing a pixel bit length of an input image based on pixel bit length increase information; transform means for transforming output data of the pixel bit length increasing means; entropy encoding means for entropy-encoding output data of the transform means; non-compression encoding means for non-compression-encoding input data; multiplexed data selection means for selecting output data of the entropy encoding means or output data of the non-compression encoding means; and multiplexing means for multiplexing the pixel bit length increase information in a bitstream, wherein a pixel bit length of an image corresponding to the output data of the entropy encoding means and a pixel bit length of an image corresponding to the output data of the non-compression encoding means are different from each other.

Abstract: When performing noise removal of an image signal by applying the background art, there is a case where a filter coefficient that causes an edge constituting the image to become blurred against expectation due to the influence of noise in the image signal is set. The methods to solve this problem include a method for calculating a wide range of image feature values needed for correction of the filter coefficient, but this method has the problem that it requires a significant amount of calculation, thus increasing the calculation cost. To solve the above problem, the present invention is provided with an image signal input means, a wavelet transformation means, a first structure/texture separating means, a texture component degenerating means, a first combining means, an inverse wavelet transformation means, and a second structure/texture separating means. The present invention is further provided with a second texture component degenerating means, a second combining means, and an image signal output means.

Abstract: An image processing method separates an input image into a skeleton component, and residual component. In the method a local variation amount, which is a variation amount between a target pixel and a pixel adjacent to the target pixel, is calculated; a skeleton component extraction filter weight is calculated based on the local variation amount; an image feature amount of a gradient direction of a pixel value around the target pixel, and an image feature amount of a gradient strength of the pixel value around the target pixel, are calculated; the skeleton component extraction filter weight is corrected based on these image feature amount; a skeleton component extraction filter coefficient is calculated based on the corrected skeleton component extraction filter weight; and the skeleton component is extracted by applying skeleton component extraction filtering to the target pixel using the calculated skeleton component extraction filter coefficient.

Abstract: The present invention provides an image processing method, an image processing device, and an image processing program which require low computing costs and with which it is possible to minimize noise amplification and halo generation caused by HDR. An image processing device according to one embodiment of the present invention has: a multi-resolution image generation means for generating a multi-resolution image; a correction amount calculation means for calculating a brightness correction amount on the basis of a lowest-resolution image of the multi-resolution image, a differential image between adjacent resolutions of the multi-resolution image, and edge information calculated at each resolution of the multi-resolution image; and a noise suppression means for calculating, on the basis of the lowest-resolution image, the differential image between adjacent resolutions, the edge information, and the brightness correction amount, an image after brightness correction in which a noise component is suppressed.

Abstract: A video decoding device for decoding video using inter prediction comprises decoding control unit setting partition type of CU to be decoded to a type other than N×N which indicates a size of PU obtained by dividing a CU to be decoded is a minimum size, when a prediction mode of the CU to be decoded is an inter prediction and a size of the CU to be decoded is equal to a minimum CU size.