Eye position detecting apparatus and method therefor

• US 6,072,892 A
• Filed: 04/03/1998
• Issued: 06/06/2000
• Est. Priority Date: 08/20/1997
• Status: Expired due to Term

• Current Assignee: Wi-LAN Inc.
• Original Assignee: Daewoo Electronics (Government of South Korea)
• Inventors: Kim, Seong-Beom
• Primary Examiner(s): Au, Amelia
• Assistant Examiner(s): Bali, Vikkram
• Application Number:
  US09/054,687
• Time in Patent Office: 795 Days
• Field of Search: 382/115, 382/116, 382/117, 382/118, 382/170, 382/172, 382/207, 351/210, 345/134, 340/825.34
• US Class Current: 382/117
• CPC Class Codes: G06T 2207/30201 Face
• CPC Class Codes: G06T 7/70 Determining position or orientation of objects or cameras camera calibration G06T7/80
• CPC Class Codes: G06V 40/18 Eye characteristics, e.g. of the iris

Abstract

An eye position detecting apparatus for detecting an eye position in a facial image utilizes a histogram thereof, wherein the facial image contains a plurality of pixels each of which is represented by a grey level value. The eye position detecting apparatus includes a window generator for windowing the facial image, a histogram extractor for extracting a histogram from the windowed facial images, a peak detector for detecting peaks of the histogram and for comparing the number of the peaks with 3, an area calculator for calculating the areas of the histogram for each of the windowed facial images when the number is equal to 3 and a selector for selecting a largest area among the calculated areas and determining the eye position by using largest area.

Claims

1. A method for detecting an eye position in a facial image by using a histogram therefor, wherein the facial image contains a plurality of pixels each of which is represented by a grey level value, the method comprising the steps of:

(a) filtering the facial image to thereby obtain a filtered facial image;

(b) generating a window within the filtered facial image;

(c) extracting a histogram from the window;

(d) detecting peaks of the histogram;

(e) comparing the number of peaks with 3, wherein if the number is equal to 3, the window is regarded as a candidate window;

(f) determining two peaks having a maximum and a minimum pixel values among the three peaks of the candidate window;

(g) calculating an A_T of the candidate window, wherein the A_T is a sum of A_MAX and A_MIN, and the A_MAX represents the pixel number in a region enclosing the peak having the maximum pixel value and the A_MIN represents the pixel number in a region enclosing the peak having the minimum pixel value;

(h) storing the A_T '"'"'s for the candidate windows and corresponding position vectors; and

(1) selecting a largest A_T among the stored A_T '"'"'s and outputting a corresponding position vector as the eye position.

2. The method of claim 1, wherein in case when both sides of eye positions are selected, position vectors corresponding to two maximum A_T '"'"'s among the A_T '"'"'s are outputted as the both sides of the eye positions.