Image analysis – Applications – Target tracking or detecting
Reexamination Certificate
2000-01-10
2003-10-28
Tran, Phuoc (Department: 2621)
Image analysis
Applications
Target tracking or detecting
C382S165000, C382S190000, C382S209000
Reexamination Certificate
active
06639998
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method of detecting a specific object in an image signal and particularly to a method of detecting a face of a character in a moving picture file.
2. Background of the Related Art
Recently, technology for digital image signal processing has been greatly developed and has been applied in various fields. For example, the digital image signal processing technology may be used in a search system for automatically editing only the face of a specific character in a moving picture file of a movie or drama, or in a security system for permitting access to only persons registered in the system. The performance of such systems basically depend on the accuracy and speed of the detecting a desired object. Accordingly, various methods of detecting a desired object have been proposed in the related art.
In “
Pedestrian Detection Using Wavelet Templates”
, CVPR97, June 17-19, MIT, a face is detected by using predetermined face templates of a plurality of characters, templates for each character ranging from a minimum size to a maximum size. When the image signal is input to the system, the frames are scanned and matched with the face templates of various size and character to detect the face. As a result, a great number of template matching may be required, thereby increasing the processing time for detecting a desired object.
In “
Automatic Extraction of Face from Color Reversal Film using Statistical Multistep Technique”
, ACCV '95 Second Asian Conference on Computer Vision, December 5-8, a method of detecting the face region of a character in a still image has been proposed. According to this face detecting method, the values expressed in the RGB (red, green, blue) color space are converted to the HSV (hue, saturation, value) color space, and the pixels representing a skin color are detected. Thereafter, segmentation of the detected skin color region is performed using an edge detection, hole filling, and gap filling operation. Thus, a final face region is detected by identifying the face region of a desired character using patterns prepared for estimating the detected region.
However, the relative accuracy of face detection is lower in comparison to other face detection methods because a still image is used which requires less information than the moving image. Also, this method utilizes only skin color information in detecting the face region, but such skin color information may changes according to the ethnic origin and circumstances such as the illumination, background, or makeup state of the character. Moreover, since the method requires the conversion from the RGB to the HSV color space, more processing time is required in detecting the face region.
Furthermore, a face region detection method using object-dependent skin color information is described in “
Extraction of facial regions features using color and shape information”
, Karin Sobotta, pp 421-425 ICPR '96. However, the face region detection using only the skin color information is not as reliable. As discussed above, since the skin color information may change according to ethnic origin as well as circumstances, it is very difficult to pre-define the range of the skin color values.
Other detection methods include automatically detecting the face region of a character if the character moves on a screen of a video camera and then identifying the character if the detected face has been registered. According to this type of face detection, the whole movement of a character is detected using image difference, and the face of the character is identified using the nervous system. Movements caught by the image difference may include movements of parts such as the body, arms or legs rather than the face, or noises belonging to the background. Thus, accuracy may not be high in detecting the face region from the image which moves as a whole. Also, in case of fast movement of a character, a significant time would be required to detect the face region.
SUMMARY OF THE INVENTION
Accordingly, an object of the present invention is to solve at least the problems and disadvantages of the related art.
An object of the present invention is to provide a method of detecting a specific object in an image signal both accurately and rapidly.
Another object of the present invention is to provide a method of detecting a face in an image signal both accurately and rapidly.
Still another object of the present invention is to provide a method of detecting a specific object or partial regions of the specific object in an image signal.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objects and advantages of the invention may be realized and attained as particularly pointed out in the appended claims.
To achieve the objects and in accordance with the purpose of the invention, as embodied and broadly described herein, the method of detecting a specific object in an image signal comprises detecting the specific object from a first input image frame using a general feature of the object, performing an adaptive study with respect to additional information which depends on the detected object, and detecting a final object from the following input image frames by applying the additional feature to the object detected using the general feature.
In another aspect of the present invention, a method of detecting a specific object in an image signal comprises detecting the specific object from an input image frame using a detection algorithm, performing a template matching n times within a limited effective range based on the detected object, and detecting a final object using the results of the template matching.
In still another aspect of the present invention, a method of detecting a specific object in an image signal comprises dividing a color range (W) of the object included in an input image signal into a plurality of level sections, detecting color images which satisfy a color range condition of the divided level sections with respect to the image frame, and detecting and confirming the object from the detected color images using a confirmation algorithm.
REFERENCES:
patent: 5912980 (1999-06-01), Hunke
patent: 6141434 (2000-10-01), Christian et al.
patent: 6263113 (2001-07-01), Abdel-Mottaleb et al.
patent: 6292575 (2001-09-01), Bortolussi et al.
Kim Hycon Jun
Lee Jin Soo
LG Electronics Inc.
Tran Phuoc
LandOfFree
Method of detecting a specific object in an image signal does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of detecting a specific object in an image signal, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of detecting a specific object in an image signal will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3161078