Image analysis – Applications – Target tracking or detecting
Reexamination Certificate
1999-01-19
2001-10-02
Mancuso, Joseph (Department: 2621)
Image analysis
Applications
Target tracking or detecting
C348S169000
Reexamination Certificate
active
06298145
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention pertains to digital image processing. More particularly, this invention relates to apparatus and a method for extracting image frames suitable for printing and/or visual presentation from compressed image data.
2. Description of the Related Art
As is known, digital or digitized video data provides significant improvements over analog video media in many aspects. Digital video data can be readily compressed and decompressed, thereby enabling efficient transmission between remote sites. Efficient compression and decompression of digital video data also enhance performance in storage and retrieval of digital video data. Digital video data also allows better video image manipulation than analog video media.
With the advances in digital storage, Internet, and multimedia computing technologies, digital video can now be stored, accessed, and manipulated more easily. For example, a video program or a collection of video programs can be stored in an optical disc and be accessed by a computer system connected to an optical disc driver with the optical disc either stationed locally or remotely. The remotely located optical disc is accessed via a network. This arrangement allows creation of a video image database that can store a collection of video programs. Each video program can be retrieved from the image database for display and/or manipulation.
However, one of the challenging problems is managing and presenting image or video information. Segmenting a video program into a number of appropriate units (i.e., clips) and characterizing each clips with key frames has been suggested as one possible mechanism for organizing and representing video information. Using this prior art approach, the key frames of a video sequence are extracted from the video sequence to represent the content of the video sequence in abstract manner. Known prior art technologies can be used to extract key frames from a video sequence. However, none of the known frame extraction technologies determines if the extracted key frame is suitable for printing or visual presentation or not (e.g., blurred or not).
Moreover, it may also be desirable for a user to select or extract a particular image frame from a video program/sequence to be printed for sharing or display. The user may also want to select and print the frames that are representative of the video program. As is known, two types of image frames are typically suitable for printing or display. The first type of image frame is an individual still image frame. Still image frames are characterized as images without motion. If a still image is not blurred, the image is suitable for printing or display. The other type of image frame is a super-resolution image that can be extracted from a video sequence. For example, a super-resolution image can be extracted from a video sequence that has undergone some global camera motion. Unlike still images, the super-resolution images are composed from images that have undergone global motion. However, like still images, a less blurred super resolution image will also offer better printing quality.
The user typically does the frame selection while watching the video program. When the user feels that a desirable frame is shown, the user can stop playing the video program (i.e., pause) and then select the desired image frame for printing. The user may also go back a number of image frames to determine which one of the image frames backwards is best suited for printing.
Disadvantages are, however, associated with this prior arrangement of selecting image frames for printing or visual presentation. One disadvantage is that the selection typically requires the user to be in a constant alert state for relatively long period of time. This typically causes the user to be tired and loose focus quickly. Afterwards, it may be very hard for the user to look even at the display. However, the user may still be required to finish the selection of the image frames from the remaining video program or sequence relatively quickly.
Another disadvantage is that the image frames visually selected by the user may not necessarily be suitable for printing. This is due to the fact that it is often difficult for the user to determine visually how blurred the image frame is relatively quickly. This may also be due to the fact that it is often difficult for the user to determine how much motion information of the current image frame with respect to adjacent frames is contained in the image frame in a relatively short period of time. It will take the user prohibitively more time if the user views the video sequence frame by frame and repeatedly in order to select the image frames that are suitable for printing or visual presentation.
Thus, there exists a need to automatically identify and select image frames that are suitable for printing and/or visual presentation from a compressed video without user intervention.
SUMMARY OF THE INVENTION
One feature of the present invention is to allow automatic selection of image frames suitable for printing and/or visual presentation.
Another feature of the present invention is to automatically identify or select image frames suitable for printing and/or visual presentation from a compressed image data.
A further feature of the present invention is to automatically identify or select image frames suitable for printing and/or visual presentation from a compressed image data by automatically extracting and analyzing motion and human face information in the image frames.
A still further feature of the present invention is to automatically identify or select image frames suitable for printing and/or visual presentation from a compressed image data by detecting the degree of blur of the image frames directly using the information contained in the compressed image data.
An image processing system for automatically extracting image frames suitable for printing and/or visual presentation from a compressed image data is described. The image processing system includes a face detector that detects if an image frame contains at least a face. The image frame is regarded as suitable for printing and/or visual presentation if the image frame is determined to contain a face.
The image processing system also includes a blur detector that determines a blur indicator value of the image frame directly using the information contained in the compressed image data if the image frame is determined to contain a face. The blur detector indicates that the image frame is suitable for printing and/or visual presentation if the blur indicator value of the image frame is less than a predetermined threshold.
Moreover, an image processing system for automatically extracting image frames suitable for printing and/or visual presentation from a compressed image data is described. The system includes a motion analyzer that detects if an image frame contains motion with respect to its adjacent image frames. The image frame is regarded as suitable for printing and/or visual presentation if the motion analyzer determines that the image frame does not contain motion.
The image processing system also includes a blur detector that is coupled to the motion analyzer to determine the blur indicator value of the image frame if the image frame is determined not to contain motion. The blur detector indicates that the image frame is suitable for printing and/or visual presentation if the blur indicator value of the image frame is less than a predetermined threshold.
Furthermore, an image blur detector that detects if an image frame of a compressed image data is blurred is also described. The blur detector includes an extractor that extracts DCT (Discrete Cosine Transform) coefficients directly from the DCT compressed image. The blur detector also includes a detection module that computes a blur indicator value of the image frame by dividing energy of the high frequency DCT coefficients with that of all non-zero DCT coefficients of the image frame. This makes the blur indicator value of the ima
Marichal Xavier
Zhang Hong Jiang
Hewlett--Packard Company
Li Thomas X.
Mancuso Joseph
Tabatabai A
LandOfFree
Extracting image frames suitable for printing and visual... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Extracting image frames suitable for printing and visual..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Extracting image frames suitable for printing and visual... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2561346