Image processing method and apparatus

Patent metadata

Field	Value
Publication number	US-2018060997-A1
Application number	US-201715429721-A
Country	US
Kind code	A1
Filing date	Feb 10, 2017
Priority date	Aug 31, 2016
Publication date	Mar 1, 2018
Grant date	—

Abstract

An image processing apparatus configured to perform a two-dimensional (2D) fast Fourier transform (FFT) with respect to image data includes a first core and a second core, each of the first core and the second core including a plurality of processors configured to perform a one-dimensional (1D) FFT; and a controller configured to control the first core and the second core to perform a primary two-dimensional (2D) FFT and a secondary 2D FFT with respect to the image data by repeatedly performing the 1D FFT.

First claim

What is claimed is:

1 . An image processing apparatus configured to perform a two-dimensional (2D) fast Fourier transform (FFT) with respect to image data, the image processing apparatus comprising:
a first core and a second core, each of the first core and the second core comprising a plurality of processors configured to perform a one-dimensional (1D) FFT; a memory configured to store data output from the first core and the second core; and a controller configured to control the first core and the second core to perform a primary 2D FFT and a secondary 2D FFT with respect to the image data by repeatedly performing the 1D FFT.

2 . The image processing apparatus of claim 1 , further comprising a plurality of operators configured to perform a focus term operation and a depth summation operation with respect to primary 2D FFT-transformed image data generated by the primary 2D FFT, to generate depth-summed image data,
wherein the controller is configured to input the depth-summed image data to the second core.

3 . The image processing apparatus of claim 1 , further comprising:
a plurality of operators configured to perform a focus term operation and a depth summation operation with respect to the primary 2D FFT-transformed image data generated by the primary 2D FFT, to generate depth-summed image data, wherein the controller is configured to store the depth-summed image data in the memory, read the depth-summed data from the memory, and input the read data to the first core.

4 . The image processing apparatus of claim 1 , wherein the first core is configured to alternately perform the primary 2D FFT with respect to a current frame of the image data and the secondary 2D FFT with respect to the current frame.

5 . The image processing apparatus of claim 1 , wherein the second core is configured to alternately perform the primary 2D FFT with respect to a current frame of the image data and the secondary 2D FFT with respect to the current frame.

6 . The image processing apparatus of claim 1 , wherein the first core is configured to perform the 1D FFT in a row direction, and the second core is configured to perform the 1D FFT in a column direction.

7 . The image processing apparatus of claim 1 , wherein the first core is configured to perform the 1D FFT in a column direction, and the second core is configured to perform the 1D FFT in a row direction.

8 . An image processing method of performing two-dimensional (2D) fast Fourier transform (FFT) with respect to image data, the image processing method comprising:
performing, by a first core and a second core, a primary 2D FFT with respect to the image data to generate primary 2D FFT-transformed data; performing a focus term operation and a depth summation operation with respect to the primary 2D FFT-transformed data to generate depth-summed data; and performing, by the first core and the second core, a secondary 2D FFT with respect to the depth-summed data.

9 . The image processing method of claim 8 , wherein the performing of the 2D FFT comprises:
inputting the depth-summed data to the second core, and performing, by the second core, a one-dimensional (1D FFT with respect to the depth-summed data to generate 1D FFT-transformed data; storing the 1D FFT-transformed data generated by the second core in a memory; and reading the stored data from the memory, inputting the read data to the first core, and performing, by the first core, the 1D FFT with respect to the input data.

10 . The image processing method of claim 8 , wherein the performing of the 2D FFT comprises:
storing the depth-summed data in a memory; reading the depth-summed data from the memory, inputting the read data to the first core, and performing, by the first core, a 1D FFT with respect to the input data to generate 1D FFT-transformed data; storing the 1D FFT-transformed data generated by the first core in a memory; and reading the stored data from the memory, inputting the read data to the second core, and performing, by the second core, the 1D FFT with respect to the input data.

11 . The image processing method of claim 8 , further comprising alternately performing, by the first core, the primary 2D FFT with respect to a current frame of the image data and the secondary 2D FFT with respect to the current frame.

12 . The image processing method of claim 8 , further comprising alternately performing, by the second core, the primary 2D FFT with respect to a current frame of the image data and the secondary 2D FFT with respect to the current frame.

13 . The image processing method of claim 8 , wherein the first core performs a one-dimensional (1D) FFT in a row direction, and the second core performs the 1D FFT in a column direction.

14 . The image processing method of claim 8 , wherein the first core performs a one-dimensional (1D) FFT in a column direction, and the second core performs the 1D FFT in a row direction.

15 . A non-transitory computer-readable recording medium having recorded thereon a program which, when executed, causes an image processing apparatus to perform an image processing method comprising:
performing, by a first core and a second core, a primary two-dimensional (2D) fast Fourier transform (FFT) with respect to image data to generate primary 2D FFT-transformed data; performing a focus term operation and a depth summation operation with respect to the primary 2D FFT-transformed data to generate depth-summed data; and performing, by the first core and the second core, a secondary 2D FFT with respect to the depth-summed data.

16 . An image processing apparatus comprising:
a first core configured to perform a first operation of one-dimensional (1D) FFT with respect to a frame of image data and thereby generate intermediate data; and a second core configured to perform a second operation of 1D FFT with respect to the intermediate data and thereby generate holographic data, wherein the first and second operations operate in perpendicular directions with respect to the frame of the image data.

17 . The image processing apparatus of claim 16 , further comprising a memory configured to store the intermediate data.

18 . The image processing apparatus of claim 16 , wherein the image data comprises a first frame which is the frame and a second frame adjacent to the first frame, and the first core is further configured to process image data of the first frame and the second frame based on time division.

19 . The image processing apparatus of claim 16 , wherein the frame comprises n lines, and the second core comprises n processors configured to simultaneously perform the second operation of 1D FFT on the respective n lines.

20 . The image processing apparatus of claim 16 , further comprising a camera configured to capture the frame of the image data.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Kim Hojung
• Kim Yongkyu
• Lee Hongseok

Classifications

• G06T1/20 Primary

  Processor architectures; Processor configuration, e.g. pipelining · CPC title

• G06T2200/28

  involving image processing hardware · CPC title

• G06T2207/10004

  Still image; Photographic image · CPC title

• G06T2207/20056

  Discrete and fast Fourier transform, [DFT, FFT] · CPC title

• G06T2207/10028

  Range image; Depth image; 3D point clouds · CPC title