ASTIGMATIC DEPTH FROM DEFOCUS IMAGING

• US 20150015676A1
• Filed: 04/18/2014
• Published: 01/15/2015
• Est. Priority Date: 04/18/2013
• Status: Active Grant

• Current Assignee: Ellis Amalgamated LLC
• Original Assignee: Ellis Amalgamated LLC
• Inventors: Favalora, Gregg E., Tsukanov, Vitaly G., Bashtyk, Yuriy V.
• Granted Patent: US 9,532,032 B2
• Time in Patent Office: Days
• US Class Current: 348/49
• CPC Class Codes: G06T 2200/04 involving 3D image data
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• CPC Class Codes: H04N 13/254 in combination with electromagnetic radiation sources for illuminating objects
• CPC Class Codes: H04N 13/271 wherein the generated image signals comprise depth maps or disparity maps

Abstract

A rangefinder has a depth camera and a computer. The depth camera is positioned to receive electromagnetic energy from a scene. Imaging optics and a sensor in the camera generate an input image based on electromagnetic energy. The computer receives the input image, calculates intermediate images by applying differentiation operators along different axes to the input image, creates a merit function map, and computes a depth map from the merit function map.

Claims

1. An apparatus comprising a rangefinder, said rangefinder comprising a depth camera and a computer, wherein said depth camera is positioned to receive electromagnetic energy from a scene, wherein said depth camera comprises imaging optics and a sensor configured to generate an input image based on said electromagnetic energy, and wherein said computer is programmed to receive said input image, to calculate a first intermediate image from said input image, to calculate a second intermediate image from said input image, to create a merit function map based on said first and second intermediate images, and to compute a depth map from the merit function map.

2. The apparatus of claim 1, wherein said imaging optics are astigmatic.

3. The apparatus of claim 1, wherein said computer is programmed to calculate said first intermediate image by applying a first differentiation operator to said input image, and wherein said computer is programmed to calculate said second intermediate image by applying a second differentiation operator to said input image, wherein said first and second differentiation operators are along different axes.

4. The apparatus of claim 1, wherein said computer is programmed to calculate said first intermediate image and said second intermediate image by single-coordinate differentiation of said input image in the spatial domain.

5. The apparatus of claim 1, wherein said computer is programmed to calculate said first intermediate image and said second intermediate image by anisotropic filtering of said input image in the frequency domain.

6. The apparatus of claim 1, wherein said computer is programmed to create said merit function map in a way that includes blurring said intermediate images.

7. The apparatus of claim 1, wherein said computer is programmed to create said merit function in a way that includes evaluation of a function of said intermediate images on a pixel-by-pixel basis.

8. The apparatus of claim 1, wherein said computer is programmed to select said first and second axes to be horizontal and vertical axes.

9. The apparatus of claim 1, wherein said computer is programmed to select said first and second axes to be orthogonal axes.

10. An apparatus comprising a computer and means for relaying information indicative of a scene to said computer, wherein said computer is programmed to receive an input image from said means for relaying information, to calculate a first intermediate image from said input image, to calculate a second intermediate image from said input image, to create a merit function map based on said first and second intermediate images, and to compute a depth map from the merit function map.

11. A method of generating a depth map, said method comprising receiving an input image indicative of a scene, calculating a first intermediate image along a first axis from said input image, calculating a second intermediate image along a second axis from said input image, said second axis differing from said first axis, creating a merit function map based on said first and second intermediate images, and computing said depth map from said merit function map.

12. The method of claim 11, wherein receiving said image comprises receiving an astigmatic image.

13. The method of claim 11, wherein calculating a first intermediate image from said input image comprises applying a first differentiation operator to said input image, and wherein calculating a second intermediate image from said input image comprises applying a second differentiation operator to said input image

14. The method of claim 11, wherein calculating a first intermediate image from said input image comprises single-coordinate differentiation of said input image in the spatial domain, and wherein calculating a second intermediate image from said input image comprises single-coordinate differentiation of said input image in the spatial domain.

15. The method of claim 11, wherein calculating a first intermediate image from said input image comprises anisotropic filtering in the frequency domain, and wherein calculating a second intermediate image from said input image comprises anisotropic filtering of said input image in the frequency domain

16. The method of claim 11, wherein creating said merit function map comprises blurring said intermediate images.

17. The method of claim 11, wherein creating said merit function comprises evaluating a function of said intermediate images on a pixel-by-pixel basis.

18. The method of claim 11, further comprising selecting said first and second axes to be horizontal and vertical axes.

19. The method of claim 11, further comprising selecting said first and second axes to be orthogonal axes.

20. The method of claim 11, creating a merit function map based on said first and second intermediate images comprises creating said merit function by pixel-by-pixel calculation using the formula