Detection of transitions between text and non-text frames in a video stream

• US 8,989,499 B2
• Filed: 10/20/2010
• Issued: 03/24/2015
• Est. Priority Date: 10/20/2010
• Status: Active Grant

• Current Assignee: Comcast Cable Communications LLC (Comcast Corporation)
• Original Assignee: Comcast Cable Communications LLC (Comcast Corporation)
• Inventors: Jojic, Oliver, Houghton, David F.
• Primary Examiner(s): DULANEY, KATHLEEN YUAN
• Application Number:
  US12/908,048
• Publication Number:
  US 20120099795A1
• Time in Patent Office: 1,616 Days
• Field of Search: 382/182, 382/176, 382/190, 725/39, 386/241, 348/700
• US Class Current: 382/190
• CPC Class Codes: G06V 20/635 Overlay text, e.g. embedded captions in a TV program
• CPC Class Codes: G11B 27/034 on discs G11B27/036, G11B27/038 take precedence
• CPC Class Codes: G11B 27/036 Insert-editing
• CPC Class Codes: H04N 21/4316 for displaying supplemental content in a region of the screen, e.g. an advertisement in a separate window
• CPC Class Codes: H04N 21/4334 Recording operations recording of a television signal H04N5/76; arrangements for recording or accumulating broadcast information or broadcast-related information H04H60/27
• CPC Class Codes: H04N 21/44008 involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59
• CPC Class Codes: H04N 21/4583 Automatically resolving scheduling conflicts, e.g. when a recording by reservation has been programmed for two programs in the same time slot
• CPC Class Codes: H04N 21/47214 for content reservation or setting reminders; for requesting event notification, e.g. of sport results or stock market stock exchange data over packet-switching network H04L12/1804; push services over packet-switching network H04L12/1859; notification of incoming messages in packet switching networks H04L51/224
• CPC Class Codes: H04N 21/488 Data services, e.g. news ticker systems specially adapted for using meteorological information in broadcast systems H04H60/71
• CPC Class Codes: H04N 21/812 involving advertisement data advertising per se G06Q30/02
• CPC Class Codes: H04N 5/445 for displaying additional information H04N5/50 takes precedence
• CPC Class Codes: H04N 5/76 Television signal recording

Abstract

Detecting the start of a credit roll within video program may allow for the automatic extension of video recordings among other functions. The start of the credit roll may be detected by determining the number of text blocks within a sequence of frames and identifying a point in the sequence of frames where a difference between the number of text blocks in frames occurring before the point and the number of text blocks in frames occurring after the point is greatest and exceeds a specified threshold. Text blocks may be identified within each frame by partitioning the frame into one or more segments and recording the segments having a pixel of a sufficiently high contrast. Contiguous segments may be merged or combined into single blocks, which may then be filtered to remove noise and false positives. Additional content may be inserted into the credit roll frames.

Claims

1. A method comprising:

determining, by a computing device, a boundary between a first portion of a video and a second portion of the video, the determining comprising;

determining a text detection window, wherein a height of the text detection window is less than a width of the text detection window if a text language of the video is intended to be read in a horizontal manner,

determining, using the text detection window, a number of text blocks within a first plurality of video frames of the video and a number of text blocks within a second plurality of video frames following the first plurality of video frames in the video, and

comparing the number of text blocks within the first plurality of video frames with the number of text blocks within the second plurality of video frames.

2. The method of claim 1, wherein determining the text detection window comprises defining the width of the text detection window to be less than the height of the text detection window if the text language of the video is intended to be read in a vertical manner.

3. The method of claim 1, further comprising:

determining a density of pixels, in a text block of the first plurality of video frames, having a contrast greater than a contrast threshold; and

determining that the text block is a false positive in response to determining that the determined density of the pixels in the text block is lower than a contrast density threshold.

4. The method of claim 3, wherein the contrast density threshold is greater than a width of the text block divided by the width of the text detection window.

5. The method of claim 1, further comprising automatically extending a recordation of the video until at least a time corresponding to the determined boundary.

6. The method of claim 1, further comprising inserting content different from the video into the video at a point past the determined boundary.

7. The method of claim 6, wherein the inserted content overlays the second portion and comprises an advertisement.

8. The method of claim 1, further comprising filtering out one or more determined sections that do not meet at least one of:

a height threshold and a width threshold.

9. A method comprising:

dividing, by a computing device, a video frame, of a plurality of video frames in a video, into sections using a text detection window of a defined size;

determining, within the sections of the video frame, contiguous sections having at least a portion with a contrast greater than a contrast threshold;

combining the determined contiguous sections to form one or more text blocks;

determining a number of formed text blocks in each of the plurality of video frames; and

applying a wavelet to the determined number of formed text blocks to determine a beginning of a credit roll in the video.

10. The method of claim 9, further comprising filtering the one or more formed text blocks, the filtering comprising:

determining a density of pixels having a contrast above the contrast threshold within a text block of the one or more formed text blocks; and

removing, from consideration, the text block of the one or more formed text blocks in response to determining that the determined density is lower than a density threshold.

11. The method of claim 10, wherein the density threshold is greater than a size of the text block divided by the size of the text detection window.

12. The method of claim 9, wherein applying the wavelet comprises:

positioning the wavelet at a first point in the plurality of video frames;

determining a negative sum of a number of formed text blocks determined for each of a first set of video frames that are chronologically prior to the first point;

determining a positive sum of a number of formed text blocks determined for each of a second set of video frames that are chronologically after the first point; and

determining a difference between the positive sum and the negative sum.

13. The method of claim 12, wherein applying the wavelet is performed for multiple points in the plurality of video frames and wherein the beginning of the credit roll corresponds to a point having a greatest difference between the positive sum and the negative sum.

14. A method comprising:

determining a number of text blocks within each of a plurality of video frames of a video content item, the determining comprising;

dividing each of the plurality of video frames into sections based on a size of a text detection window, wherein a relationship between a height and a width of the text detection window is defined based on an intended reading direction of text included in the video content item, and

determining which of the sections comprises a portion having a determined contrast greater than a contrast threshold; and

determining a boundary between a first portion and a second portion of the video content item based on the determined number of text blocks within each of the plurality of video frames.

15. The method of claim 14, further comprising defining the height of the text detection window to be less than the width of the text detection window when the intended reading direction is horizontal.

16. The method of claim 14, further comprising defining the width of the text detection window to be less than the height of the text detection window when the intended reading direction is vertical.

17. A method comprising:

determining, by a computing device, text areas within a first plurality of video frames of a content item, the determining comprising dividing each of the first plurality of video frames into sections based on a size of a text detection window comprising a height and a width, the height and the width being defined by an intended reading direction of text in the video frames, and evaluating a contrast of each of the sections; and

determining whether a boundary between a first portion and a second portion of the content item exists within the first plurality of video frames based on a number of text areas within each of the first plurality of video frames.

18. The method of claim 17, wherein evaluating the contrast of each of the sections comprises comparing the contrast with a contrast threshold.

19. The method of claim 17, further comprising determining text areas within a second plurality of video frames different from the first plurality of video frames in response to determining that the boundary between the first portion and the second portion does not exist within the first plurality of video frames.