Patents by Inventor Tadayoshi Nakayama
Tadayoshi Nakayama has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 7236997Abstract: The present invention is applicable to a wavelet transform in JPEG 2000 and makes it possible to implement a convolution operation while reversibility is achieved. To this end, an implementation of a two-stage integer-type lifting operation shown in FIG. 28A is decomposed in the order of FIGS. 28B and 28C, and finally is separated into a real-number processing unit and integerizing processing unit as shown in FIG. 28C.Type: GrantFiled: October 24, 2002Date of Patent: June 26, 2007Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Patent number: 7227998Abstract: This invention reliably encodes an image while continuously inputting the image with a relatively simple configuration. For this purpose, in this invention, input image data is encoded by an encoder (102) and stored in first and second memories. An encoding sequence controller (108) monitors the code amount. Upon determining that the code amount has reached a set value, the encoding sequence controller discards the data in the first memory and instructs the encoder (102) to increase the quantization step and continue encoding. Preceding encoded data is stored in the second memory. The encoded data is re-encoded by a re-encoder (109) using the same quantization step as that of the encoder (102) after the parameter is changed. The re-encoded data is stored in the first and second memories. The quantization steps set in the encoder (102) and re-encoder (109) at this time have such values that re-encoding by the re-encoder (109) is ended before time when the code amount reaches the set value again.Type: GrantFiled: June 11, 2003Date of Patent: June 5, 2007Assignee: Canon Kabushiki KaishaInventors: Tadayoshi Nakayama, Naoki Ito
-
Patent number: 7194140Abstract: This invention has as its object to provide an image processing apparatus, which can obviate the need for image re-input, can effectively generate encoded data that falls within a set size, and can minimize deterioration of image quality. To this end, an image processing apparatus according to this invention includes first discrimination unit for discriminating the type of image of each of a plurality of regions, which form image data input, second discrimination unit for discriminating if the image data is inputted by a continuous scan of images, selection unit for selecting an encoding method used in compression of each region on the basis of discrimination results of the first and second discrimination unit, first compression unit for compressing image data of each region using the encoding method selected by the selection unit, and second compression unit for compressing information that pertains to the type of image of each region.Type: GrantFiled: November 4, 2002Date of Patent: March 20, 2007Assignee: Canon Kabushiki KaishaInventors: Naoki Ito, Tadayoshi Nakayama, Ken-ichi Ohta, Hidefumi Osawa, Shinichi Kato
-
Patent number: 7188132Abstract: The present invention provides a method and apparatus that transforms input signals X0, X1, X2, X3 using a 4-point Hadamard transform matrix, and of the resulting transformation data, rounds up the LSB of the first transformed data and rounds down the LSB of the remaining odd-numbered data. When restoring the data to its original state the rounding is executed after a Hadamard inverse transformation is performed, thereby restoring the data to its original state X0, X1, X2, X3.Type: GrantFiled: December 19, 2002Date of Patent: March 6, 2007Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Publication number: 20070025632Abstract: This invention allows Huffman encoding using a common Huffman table according to basic quantization values Qi,j even when image data expressed by n bits falling within the range from L to K is JPEG-coded, and can suppress the Huffman table size from increasing. To this end, a basic quantization table storage unit stores quantization step values Q0,0 to Q7,7 used in baseline JPEG coding. A minimum quantization step generator outputs a minimum quantization step value Qn—min to a comparator/selector according to the number n of bits of each color component of image data to be coded. The comparator/selector compares the quantization step values Q0,0 to Q7,7 with Qn—min and selects larger ones, and outputs the comparison results to a quantizer as Q?i,j. The quantizer stores Q?i,j in a quantization table storage unit and quantizes orthogonal transformation coefficients output from a DCT transformer.Type: ApplicationFiled: July 26, 2006Publication date: February 1, 2007Applicant: CANON KABUSHIKI KAISHAInventor: Tadayoshi Nakayama
-
Patent number: 7127115Abstract: The present invention allows an image to be coded within a target size without necessitating the image to be input again during the coding of the image, with a mode reflecting a user's intention for coding. To solve this problem, input image data is coded at coding unit 102 and stored into first and second memories, respectively. Coding sequence unit 108 monitors the quantity of codes. When a set value is determined to be reached, coding sequence unit 108 makes data in first memory to be discarded and directs coding means to further increase a quantization step, and continues coding. As previous coded data is stored in second memory, the data is re-coded with the same quantization step as that of coding unit 102 after changing of a parameter at re-coding unit 109, and the re-coded data is stored into first and second memory.Type: GrantFiled: March 21, 2003Date of Patent: October 24, 2006Assignee: Canon Kabushiki KaishaInventors: Hidefumi Osawa, Tadayoshi Nakayama, Ken-ichi Ohta, Shinichi Kato, Naoki Ito
-
Patent number: 7106911Abstract: An image input through an input unit is compressed by an encoding unit and stored in first and second memories. A first counter counts the data in its code amount. When the amount of encoded data generated reaches a predetermined size, an encoding sequence control unit sets quantization steps in the encoding unit and re-encoding unit to increase compression ratios. The encoding sequence control unit clears the first memory, causes the re-encoding unit to re-encode the encoded data stored in the second memory, and stores the resultant data in the first memory. Since the encoding unit continues encoding data with the set quantization step, the encoded data are stored in the first memory from the start of the image. Subsequently, every time the data amount stored in the first memory reaches a predetermined amount, the quantization step is increased, and the processing is repeated.Type: GrantFiled: June 17, 2002Date of Patent: September 12, 2006Assignee: Canon Kabushiki KaishaInventors: Ken-ichi Ohta, Tadayoshi Nakayama, Hidefumi Osawa, Shinichi Kato, Naoki Ito
-
Publication number: 20060045369Abstract: A data transform method comprises a linear transform step of acquiring 16 linear Hadamard transform coefficients by multiplying the 16 pieces of data by an Hadamard transform matrix, an offset step of classifying the 16 linear Hadamard transform coefficients into four groups, each group containing an odd number of coefficients, and adding a predetermined offset value to linear Hadamard transform coefficients in every group, and an integral step of acquiring the lossless Hadamard transform coefficients by truncating decimal fractions down from a decimal point from the linear Hadamard transform coefficients added said predetermined offset value. With this configuration, a 16-point lossless Hadamard transform is implemented using only one rounding process.Type: ApplicationFiled: August 18, 2005Publication date: March 2, 2006Applicant: CANON KABUSHIKI KAISHAInventor: Tadayoshi Nakayama
-
Publication number: 20060039626Abstract: The present invention performs a lossless four-point orthogonal transformation with reduced rounding errors using a simple configuration.Type: ApplicationFiled: August 17, 2005Publication date: February 23, 2006Applicant: CANON KABUSHIKI KAISHAInventor: Tadayoshi Nakayama
-
Patent number: 6996593Abstract: This invention has as its object to suppress an increase in circuit scale and to simplify a circuit structure by executing a filter process using a plurality of arithmetic units each of which makes multiplication and addition. To achieve this object, image data Yn+2, Yn+3, and Yn+4 to be processed are read out, and three lattice point data d?n+1, S?n, and dn?1 are respectively read out from sequences H1, H2, and H3 corresponding to line buffers that store the lattice point data. d?n+3=Yn+3+?(Yn+2+Yn+4) is computed, and d?n+3 is stored in the sequence H1. S?n+2+?(d?n+1+d?n+3) is computed, and S?n+2 is stored in the sequence H2. d?n+1=d?n+1+y(S?n+2+S?n) is computed, and dn+1 is stored in the sequence H3. Sn=S?n+?(dn?1+dn+1) is computed, and Sn and dn+1 are output to the next processing stage.Type: GrantFiled: October 22, 2001Date of Patent: February 7, 2006Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Publication number: 20050276500Abstract: This invention generates target encoded data by suppressing an arithmetic precision drop by executing processes such as orthogonal transformation and the like using the number of bits of an input image in place of reducing the number of bits at the time of input upon encoding an image. Upon generating baseline JPEG encoded data, a quantization table for an 8-bit image is stored in a quantization table storage unit. When an interpreter outputs information indicating that each color component per pixel of input image data is 16 bits, a bit shift unit multiplies the quantization table stored in the quantization table storage unit by the 8th power of 2 or {8th power of 2+1}. A quantizer quantizes coefficients output from a DCT unit on the basis of the quantization table stored in the quantization table storage unit, and a Huffman encoder encodes the quantization result to Huffman codes.Type: ApplicationFiled: June 15, 2005Publication date: December 15, 2005Applicant: CANON KABUSHIKI KAISHAInventors: Tadayoshi Nakayama, Hiroki Kishi
-
Publication number: 20050276501Abstract: This invention improves the arithmetic precision even for image data in which each component of one pixel is expressed by 8 bits by fully utilizing the 12-bit data processing performance of an Extended sequential DCT-based JPEG decoding/encoding apparatus, so that image deterioration due to JPEG compression which is observed in an image portion where the gray levels change slowly, i.e., a pseudo edge can be hardly generated, thus improving the image quality. To this end, a header interpreter interprets the header of encoded data to be decoded to determine if the encoded data is that of 8-bit image data per component or that of 12-bit image data per component, and outputs the result to a bit shift unit, rounding processor, and inverse quantizer.Type: ApplicationFiled: June 15, 2005Publication date: December 15, 2005Applicant: CANON KABUSHIKI KAISHAInventors: Tadayoshi Nakayama, Hiroki Kishi
-
Patent number: 6952501Abstract: This invention provides an image processing apparatus for executing padding processing at a high speed. For this purpose, the image processing apparatus of this invention is an image processing apparatus for processing image data on the basis of shape information, which includes a left propagation processing section (121) for propagating, of a plurality of pixel data which construct one-dimensional image data, pixel data specified by shape information to the left, a right propagation processing section (123) for propagating, of the plurality of pixel data which construct the one-dimensional image data, the pixel data specified by the shape information to the right, and a calculator group (131) for calculating the average between an output from the left propagation processing section (121) and an output from the right propagation processing section (123) to generate output pixel data.Type: GrantFiled: February 26, 2001Date of Patent: October 4, 2005Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Patent number: 6904177Abstract: A two-dimensional wavelet transform processing apparatus, more effectively utilizing hardware resource, is realized with reduced hardware construction. For this purpose, the filter processing apparatus has a vertical DWT processor (901) for performing filter processing on image data and outputting 2 types of data obtained by the processing as 1 pair of data, a rotation unit (903) for rearranging the data outputted from the vertical DWT processor by rotating the data by 90° by 2 pairs and outputting the data, and a horizontal DWT processor (905) for performing filter processing on the image data rearranged by the rotation unit and outputting 2 types of data obtained by the processing as 1 pair of data.Type: GrantFiled: March 8, 2002Date of Patent: June 7, 2005Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Patent number: 6865299Abstract: A method of performing variable-length coding for a multilevel image with simple processing, and a processing apparatus for executing this variable-length coding method at a high speed have not been established yet. According to this invention, discrete wavelet transformation processing is performed for multilevel image data to quantize the image data in units of subblocks of respective frequency components. An optimal shift parameter is easily determined at a high speed for each subblock. This realizes high-speed variable-length coding based on the parameter.Type: GrantFiled: July 18, 2000Date of Patent: March 8, 2005Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Publication number: 20040258320Abstract: A data transform processing apparatus comprising a first lossless transform circuit to perform two step ladder operation processings of receiving unweighted normalized data then outputting weighted nonnormalized rotation-transformed data, and a second lossless transform circuit to perform two step ladder operation processings of receiving the weighted nonnormalized rotation-transformed data from the first lossless transform circuit then performing inverse weighting and outputting unweighted normalized rotation-transformed data, wherein the outputs from the first lossless transform circuit are interchanged and supplied to the second lossless transform circuit.Type: ApplicationFiled: June 21, 2004Publication date: December 23, 2004Applicant: CANON KABUSHIKI KAISHAInventor: Tadayoshi Nakayama
-
Patent number: 6711295Abstract: The primary objective of the present invention is the efficient performance of Golomb-Rice encoding. To achieve this objective, an encoding method comprises the steps of receiving image data for each predetermined unit, separately employing K types of k parameters to perform Golomb-Rice coding for the image data for each predetermined unit, and generating K types of Golomb-Rice coded data, holding the K types of Golomb-Rice coded data obtained at the generation step, based on the image data, selecting one of the k parameters for each predetermined unit, and employing the selected k parameter to select and output one of the K types of Golomb-Rice coded data that are being held.Type: GrantFiled: October 5, 1999Date of Patent: March 23, 2004Assignee: Canon Kabushiki KaishaInventors: Tadayoshi Nakayama, Hiroshi Kajiwara, Takeshi Yamazaki
-
Patent number: 6690825Abstract: An image processing apparatus and method in which voice information is input to specify an object contained in a moving picture, the input voice information is recognized and analyzed, parameters representing the object to be extracted are generated based upon the result of recognition and analysis, and the object is extracted from the moving picture based upon the extracted/generated parameters.Type: GrantFiled: May 23, 2000Date of Patent: February 10, 2004Assignee: Canon Kabushiki KaishaInventor: Tadayoshi Nakayama
-
Publication number: 20040006581Abstract: Input signals are transformed with an Hadamard transformation matrix in each of the four four-point Hadamard transformation units, wherein a rounding unit rounds up the least significant bit of each of the odd number of coefficients and discards the least significant bit of each of the remaining odd number of coefficients among the four transformed coefficients output from each of the four-point Hadamard transformation units to produce four sets of four integer coefficients, and one integer coefficient is selected from each set, and four selected integer coefficients including odd number of rounded up are input to an Hadamard transformation unit and are Hadamard transformed, and the Hadamard transformed coefficients are rounded up to produce integer.Type: ApplicationFiled: June 26, 2003Publication date: January 8, 2004Applicant: CANON KABUSHIKI KAISHAInventor: Tadayoshi Nakayama
-
Publication number: 20030228063Abstract: This invention reliably encodes an image while continuously inputting the image with a relatively simple configuration. For this purpose, in this invention, input image data is encoded by an encoder (102) and stored in first and second memories. An encoding sequence controller (108) monitors the code amount. Upon determining that the code amount has reached a set value, the encoding sequence controller discards the data in the first memory and instructs the encoder (102) to increase the quantization step and continue encoding. Preceding encoded data is stored in the second memory. The encoded data is re-encoded by a re-encoder (109) using the same quantization step as that of the encoder (102) after the parameter is changed. The re-encoded data is stored in the first and second memories. The quantization steps set in the encoder (102) and re-encoder (109) at this time have such values that re-encoding by the re-encoder (109) is ended before time when the code amount reaches the set value again.Type: ApplicationFiled: June 11, 2003Publication date: December 11, 2003Applicant: CANON KABUSHIKI KAISHAInventors: Tadayoshi Nakayama, Naoki Ito