Sony Corporation

Yukihiko Mogi

Kazuhiko Nishibori

Makoto Kondo

Koji Aoyama

Takaya Hoshino

Masato Kamata

Masuyoshi Kurokawa

Nobuo Ueki

Ikuo Someya

Toshio Sarugaku

Yuki Kawaguchi

Ken Tamayama

Homare Nishizaki

An arithmetic unit which includes: a data supply section which supplies floating-point type object data to which a sign is to be added and condition data which includes a condition under which the sign is added; a sign data generating section which extracts the condition included in the condition data and generates sign data for adding the sign to the object data on the basis of the extracted condition; and an integer arithmetic operation section which performs an integer arithmetic operation while treating the object data as integer type data so as to add the sign to the object data on the basis of the sign data and the object data.

Arithmetic unit, processor, compiler and compiling method

A computation apparatus includes: a range calculation section for calculating a range of an input value that can give a predetermined discrete value obtained by discretizing a computation result of a nonlinear operation; and a discrete value output section for outputting, when the input value is input, the predetermined discrete value corresponding to the range in which the input value that has been input is contained.

Computation apparatus and method, quantization apparatus and method, audio encoding apparatus and method, and program

A mantissa/exponent splitter splits an input value X=(1+X/2)&#xd7;(2^X) into a mantissa Xand an exponent X. An interpolation processor references the mantissa/exponent splitter using the mantissa Xand determines a power value (log(1+X/2)) through an interpolation process. A logarithmic calculator determines a logarithmic value Z=logX=Y(X+log(1+X/2)) from the exponent Xand the power value from the interpolation processor. The integer/fraction splitter splits the logarithmic value Z into an integer Zand a fraction Z. The interpolation processor references a power of fraction table storage unit in response to the fraction Zand determines a power value (2^Z) through the interpolation process. The power calculator determines X=2^Z=(2^Z)&#xd7;(2^Z), thereby resulting in the input value X to the power of Y.

Apparatus, method, and program for arithmetic processing

A computation apparatus includes an inverse conversion table creation unit configured to create an inverse conversion table in which discrete values obtained by applying a predetermined conversion on predetermined data correspond to inverse conversion values obtained by applying a conversion inverse to the predetermined conversion on the discrete values, a range decision unit configured to decide in which range the predetermined data is included when the predetermined data is input among ranges where the inverse conversion values adjacent in the inverse conversion table are set as border values, and a discrete value decision unit configured to decide the discrete value corresponding to the inverse conversion value whose value is close to the predetermined data among the inverse conversion values serving as the border values of the range decided by the range decision unit.

Computation apparatus and method, quantization apparatus and method, and program

A computation apparatus includes a range table creation unit configured to create a range table in which a discrete value of a computation result obtained by applying a nonlinear computation on an input value corresponds to a range of the input value which may take the discrete value, and a search unit configured to search, when the input value is input, in the range table, for the range in which the input value is included and output the discrete value corresponding to the searched range.

A signal processing apparatus includes: a first low-pass filter filtering a data signal; and a convolving section convolving low frequency components of the data signal obtained by the first low-pass filter with a compensation filter for compensating for signal degradation attributable to low frequency components of impulse response obtained by filtering the impulse response with a second low-pass filter.

Apparatus, method and program for filtering data signals using low-frequency and compensating filters

A sampling rate converter able to obtain an amplitude characteristic that passes any frequency and able to achieve a high precision conversion without depending upon a cutoff frequency, having an up samplerfor inserting (U-) zero points between signals and raising a sampling frequency Fsi U-fold, a convolution processing unitincluding an FIR filter and interpolating a value by convolution with respect to output signals of the up sampler, and a linear interpolation blockfor selecting two points of samples from the output signal of the convolution processing unithaving a sampling frequency UFsi and finding the value at a required position from the linear interpolation, wherein the FIR filter has an impulse response becoming a filter coefficient, having a transmission function H(z) associated with a transmission function Z(z) of a pre-filter, and having a filter coefficient set by performing weighted approximation with respect to a desired characteristic associated with the frequency response of the pre-filter.

Sampling rate conversion device and method, and audio device

A vector motion detector is provided which corrects the quantity of a motion vector determined with the block matching technique between a reference block taking a reference pixel extracted from a current frame as its origin and a search block extracted from a reference frame and moving within or along the circumference of a search range and which is outside the search range. The vector motion detector makes the motion vector quantity correction so that the motion vector will fall within the search rang. Thus, it can provide a smooth motion of even an image that moves fast.

Motion vector detector and motion vector detecting method

The present invention provides an image signal processing apparatus and a method thereof in which each of the fields forming the unit-frame is specified, with respect to the inputted image signal, based on a difference value calculated in signal level between a detected pixel in a current field and a detected pixel at the same position in a field which comes one frame behind the current field, a motion vector for a field which comes two frames behind the current field is detected, with respect to the detected pixel in the current field, the detected pixel is shifted, with respect to the specified first field, in a direction opposite to the motion vector within the range of the detected motion vector, the detected pixel is shifted, with respect to the specified fourth field, in a direction along the motion vector, and the detected pixels is shifted, with respect to the specified second and third fields, so as to make the pixels gradually closer to the pixel position shifted with respect to the fourth field, in the consecutive order of the fields from the first field, in the direction along the motion vector or in the direction opposite to the motion vector.

Image signal processing apparatus and processing method

It is determined whether each block () including a plurality of pixels extracted from a basic field () has an edge or not, and then which state, stationary or non-stationary, each of the pixels forming together the block () determined to have an edge is. There is calculated an absolute value of a difference between a pixel determined to be non-stationary and a pixel (or) in each pixel position in a reference field () or pixels (to) included in the reference field and adjacent to the non-stationary pixel, and a correlation is found between such pixels in consideration according to the calculated difference absolute-value. A motion vector is determined by the block matching method according to results of the determination, and allocated to each of the pixels. Thus, the motion vector of each block is corrected to an accurate one of each of the pixels.

Motion vector correction circuit and method

An FIR filter and setting method of filter coefficients thereof enabling the prevention of breakdown of equi-ripple of weighted approximation error and preservation of gain of the pass band into approximately constant. The setting method includes five steps. The initial setting step is that setting of the FIR filter, setting of the band, setting of coefficients of a pre-filter, and setting of initial extreme value point. The first step is that interpolation polynomial equation is generated for interpolating amplitude characteristic from the present extreme value point of the amplitude characteristic of the frequency. The second step is that a new extreme value point is determined from the amplitude characteristic obtained from the interpolation polynomial equation that is obtained in the first step. The third step is that judgement is performed whether or not position of the extreme value is approximated within required range while repeating the first step and the second step. The fourth step is that finding of the filter coefficients is performed from the approximated amplitude characteristic obtained in the third step.

FIR filter and setting method of coefficient thereof

An image signal processing apparatus according to the present invention receives an image signal inputted thereto that is generated by subjecting a telecine-converted image to double speed conversion, in which signal one film frame is formed by four fields, identifies a first field on the basis of a difference value calculated between pixel signal levels, and shifts the position of a detected pixel in a vector direction of a motion vector such that an amount of shift is progressively increased as transition is made from the identified first field to the subsequent fields.

Image signal processing apparatus and method

An information processor used as a processor in a personal computer for example includes an inner-product operational unit which effects an SIMD type inner-product operation. The result of the inner-product operation effected in the inner-product operational unit is stored into a P register in the form of a word. A shift and shifting unit reads the result of the inner-product operation stored in the P register , shifts the result by an arbitrary number of bits and then clips it with the arbitrary number of bits, and stores the result into a Y register as a sub-word.

Growing community of inventors

Kanagawa, Japan

Yukihiko Mogi