FPGA based Passive Auto Focus System using Adaptive Thresholding

1
FPGA based Passive Auto Focus System using
Adaptive Thresholding

• SICE-ICASE International Joint Conference 2006
• Oct. 18-2 1, 2006 in Bexco, Busan, Korea
• Presenter???
• Number97662003

2
Outline

• Introduction
• Adaptive thresholding
• Focus measurement
• Implementation
• Experimental result
• Conclusion

3
Introduction

• The adaptive thresholding algorithm, which is
  used to decide the sharpness of the image.
• Proposes a focus measurement method using the
  adaptive thresholding algorithm, maximum
  sharpness searching function and auto focus lens
  driver.
• Implementation and experimental results,
  respectively.
• Discusses about future work.

4
Adaptive thresholding
Rh, Rv, are the horizontal thresholding result
and vertical thresholding result.
Fig. 1 A raw image
Fig. 2 An adaptive thresholding image
5
Focus measurement
Fig. 3 Focus measurement algorithm
6
The proposed focus measurement method

• Adaptive thresholding
• -applied to each pixel using the window-based
  image processing architecture.
• Rough focus searching
• -the number of black pixels in the focus window
  is counted as the sharpness value of the current
  position of the lens motor.
• Detail searching
• -retrieves the maximum sharpness value and its
  lens location.

7
Implementation
8
Auto focusing system
9
Experimental result

• Normal Condition
• Low-light Condition
• Low-contrast Condition
• Result Condition

10
Normal Condition
11
Low-light Condition
12
Low-contrast Condition
13
Result Condition
14
Conclusion

• This paper presents a FPGA based passive auto
  focusing system, which was implemented using
  adaptive thresholding.
• The pixel-based focusing method that we proposed
  is based on camera input and no additional
  devices.
• The proposed system is quite simple to implement
  and robust under low-light, low-contrast
  conditions.
• In future work, this process should be automated.

15
The End

• Thank you for your attentions !