# US 2014/0193070 A1 · Noise estimation apparatus and method thereof > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2014/0193070 A1 - Document ID: US20140193070A1 - Status: Published application - Kind: A1 - Application number: 13/897,001 - Filed: 2013-05-17 - Published: 2014-07-10 - Earliest claimed: 2013-01-08 - Assignees: Novatek Microelectronics Corp. - Applicants: Novatek Microelectronics Corp. - Inventors: Kuan-Kang LAI - Web page: https://patentgrep.com/patent/US20140193070A1 ## CPC - G06T 7/0002: {Inspection of images, e.g. flaw detection} - G06T 2207/10016: Video; Image sequence - G06T 2207/20021: Dividing image into blocks, subimages or windows - G06T 2207/20076: Probabilistic image processing - G06T 2207/30168: Image quality inspection - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract A noise estimation apparatus for calculating a noise estimation value of a frame of an image is provided. The noise estimation apparatus includes a distribution calculation unit, a variance calculation unit, a distribution curve generation module and a noise estimation unit. The distribution calculation unit generates a pixel distribution according to multiple pixel data of an ith block of the frame and multiple previous pixel data of the ith block of a previous frame. The variance calculation unit combines the pixel data and the previous pixel data to generate a variance value. The curve distribution generation module generates a curve distribution according to the variance value, and compares the pixel distribution with the curve distribution to generate a weighting value. The noise estimation unit outputs the noise estimation value according to the weighting value and the variance value corresponding to each of the blocks of the frame. ## Claims ### Claim 1 A noise estimation apparatus, for calculating a noise estimation value of a frame; the frame having an M number of overlapping or non-overlapping blocks, each of the blocks having a plurality of pixel data including a target pixel data, where M is a positive integer greater than 1; the noise estimation apparatus comprising: a distribution calculation unit, for calculating pixel numbers respectively corresponding to a plurality of pixel values to generate a pixel distribution according to the plurality of pixel data of an ith block of the frame and a plurality of previous pixel data of the ith block of a previous frame, wherein i is a positive integer between 1 and M; a variance calculation unit, for combining the pixel data of the ith block of the frame and the previous pixel data of the ith block of the previous frame, and correspondingly generating a variance value; a distribution curve generation module, for generating a distribution curve according to the variance value by regarding the target pixel data of the ith block of the frame as a basis, and comparing the pixel distribution with the curve distribution to correspondingly generate a weighting value; and a noise estimation unit, for outputting the noise estimation value according to the weighting value and the variance value corresponding to each of the blocks of the frame. ### Claim 2 (depends on 1) The noise estimation apparatus according to claim 1, wherein the distribution calculation unit comprises: a first calculation unit, for calculating the pixel numbers of the pixel data of the ith block of the frame respectively corresponding to the pixel values to output a first distribution; a second calculation unit, for calculating the pixel numbers of the previous pixel data of the ith block of the previous frame respectively corresponding to the pixel values to output a second distribution; and a combination unit, for combining the first distribution and the second distribution to generate the pixel distribution. ### Claim 3 (depends on 1) The noise estimation apparatus according to claim 1, wherein the variance calculation unit comprises: a computation unit, for combining the pixel data of the ith block of the frame and the previous pixel data of the ith block of the previous frame to correspondingly output a synthesis pixel data; and a variance acquisition unit, for generating the variance value according to the synthesis pixel data. ### Claim 4 (depends on 1) The noise estimation apparatus according to claim 1, wherein the distribution curve generation module comprises: a distribution curve generation unit, for looking up a look-up table (LUT) according to the variance value by regarding the target pixel data of the ith block of the frame as the basis to output the distribution curve; a difference calculation unit, for comparing the distribution curve with the pixel distribution to output a difference; and a weighting calculation unit, for correspondingly generating the weighting value according to the difference. ### Claim 5 (depends on 4) The noise estimation apparatus according to claim 4, wherein the distribution curve is a Gaussian distribution. ### Claim 6 (depends on 4) The noise estimation apparatus according to claim 4, wherein when a value of the difference is between a first threshold and a second threshold greater than the first threshold, the corresponding weighting value gets smaller as the difference becomes larger. ### Claim 7 (depends on 6) The noise estimation apparatus according to claim 6, wherein the first threshold and the second threshold are generated according to the variance value by means of dynamic thresholds. ### Claim 8 (depends on 1) The noise estimation apparatus according to claim 1, wherein the noise estimation unit comprises: a noise estimator, for performing a weighted averaging process on the weighting value and the variance value corresponding to each of the blocks of the frame to output the noise estimation value. ### Claim 9 (depends on 8) The noise estimation apparatus according to claim 8, wherein the noise estimation unit further comprises: a protection unit, for dividing the noise estimation value outputted by the noise estimator by a denominator when a sum of the weighting values of all of the blocks is lower than a threshold, and outputting a corresponding quotient as the noise estimation value; wherein, the denominator gets larger as the sum of the weighting values of all of the blocks becomes further lower than the threshold. ### Claim 10 (depends on 1) The noise estimation apparatus according to claim 1, wherein the frame has an M number of pixels each corresponding to one block, and the target pixel data is a central pixel data of the corresponding block. ### Claim 11 A noise estimation method, for calculating a noise estimation value of a frame; the frame having an M number of overlapping or non-overlapping blocks, each of the blocks having a plurality of pixel data including a target pixel data, where M is a positive integer greater than 1; the noise estimation method comprising: calculating pixel numbers respectively corresponding to a plurality of pixel values to generate a pixel distribution according to the plurality of pixel data of an ith block of the frame and a plurality of previous pixel data of the ith block of a previous frame, wherein i is a positive integer between 1 and M; combining the pixel data of the ith block of the frame and the previous pixel data of the ith block of the previous frame, and correspondingly generating a variance value; generating a distribution curve according to the variance value by regarding the target pixel data of the ith block of the frame as a basis, and comparing the pixel distribution with the curve distribution to correspondingly generate a weighting value; and outputting the noise estimation value according to the weighting value and the variance value corresponding to each of the blocks of the frame. ### Claim 12 (depends on 11) The noise estimation method according to claim 11, wherein the step of generating the pixel distribution comprises: calculating the pixel numbers of the pixel data of the ith block of the frame respectively corresponding to the pixel values to output a first distribution; calculating the pixel numbers of the previous pixel data of the ith block of the previous frame respectively corresponding to the pixel values to output a second distribution; and combining the first distribution and the second distribution to generate the pixel distribution. ### Claim 13 (depends on 11) The noise estimation method according to claim 11, wherein the step of generating the variance value comprises: combining the pixel data of the ith block of the frame and the previous pixel data of the ith block of the previous frame to correspondingly output a synthesis pixel data; and generating the variance value according to the synthesis pixel data. ### Claim 14 (depends on 11) The noise estimation method according to claim 11, wherein the step of generating the weighting value comprises: looking up a LUT according to the variance value by regarding the target pixel data of the ith block of the frame as the basis to output the distribution curve; comparing the distribution curve with the pixel distribution to output a difference; and correspondingly generating the weighting value according to the difference. ### Claim 15 (depends on 14) The noise estimation method according to claim 14, wherein the distribution curve is a Gaussian distribution. ### Claim 16 (depends on 14) The noise estimation method according to claim 14, wherein when a value of the difference is between a first threshold and a second threshold greater than the first threshold, the weighting value gets smaller as the difference becomes larger. ### Claim 17 (depends on 16) The noise estimation method according to claim 16, wherein the first threshold and the second threshold are generated according to the variance value by means of dynamic thresholds. ### Claim 18 (depends on 11) The noise estimation method according to claim 11, wherein the step of outputting the noise estimation value comprises: performing a weighted averaging process on the weighting value and the variance value corresponding to each of the blocks of the frame to output the noise estimation value. ### Claim 19 (depends on 18) The noise estimation method according to claim 18, wherein the step of outputting the noise estimation value further comprises: dividing the noise estimation value outputted by the noise estimator by a denominator when a sum of the weighting values of all of the blocks is lower than a threshold, and outputting a corresponding quotient as the noise estimation value; wherein, the denominator gets larger as the sum of the weighting values of all of the blocks becomes further lower than the threshold. ### Claim 20 (depends on 11) The noise estimation method according to claim 11, wherein the frame has an M number of pixels each corresponding to one block, and the target pixel data is a central pixel data of the corresponding block.