Image-data reduction technique

• US 4,692,806 A
• Filed: 04/08/1986
• Issued: 09/08/1987
• Est. Priority Date: 07/25/1985
• Status: Expired due to Term

• Current Assignee: RCA Corporation (General Electric Company)
• Original Assignee: RCA Corporation (General Electric Company)
• Inventors: Anderson, Charles H., Carlson, Curtis R.
• Primary Examiner(s): Groody, James J.
• Assistant Examiner(s): Kostak, Victor R.
• Application Number:
  US06/850,432
• Time in Patent Office: 518 Days
• Field of Search: 358/209, 358/125, 358/126, 358/180, 358/93, 358/108, 358/166, 358/107, 358/109, 358/21 R, 358/167, 358/133, 358/105, 358/903, 358/160, 382/47, 382/43, 382/54, 364/521, 901/47
• US Class Current: 375/240.08
• CPC Class Codes: G06T 3/403 Edge-driven scaling; Edge-based scaling
• CPC Class Codes: H04N 19/00 Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
• CPC Class Codes: H04N 19/102 characterised by the element, parameter or selection affected or controlled by the adaptive coding
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets

Abstract

A foveated electronic camera employing a spatial-frequency spectrum analyzer and data-reduction means utilizing movable spatial windows for converting a high-resolution, wide field-of-view image into a group of subspectrum band images ranging in resolution and field of view from a lowest-resolution image of the entire field of view to a highest-resolution image of a selectably positionable smallest spatial subregion of the entire field of view. Such a camera is useful in a surveillance camera system and in a robotic system, by way of examples.

Claims

1. An image data reduction method for use with an input video signal representing a relatively high-spatial-resolution, wide-field-of-view image that is comprised of a first given number of pixels, said method comprising the steps of:

first processing said input video signal to derive therefrom (1) a certain output video signal representing a derived relatively low-spatial-resolution image having a given field-of-view that corresponds to the field of the image represented by said input video signal, said image represented by said certain video signal being comprised of a second given number of pixels smaller than said first given number, and (2) at least one other output video signal representing a derived image having said given field-of-view, said image represented by said one other output video signal exhibiting a resolution equal to or less than said relatively high-spatial-resolution image represented by said input video signal but larger than said relatively low-spatial-resolution derived image represented by said certain output video signal, and said derived image represented by said one other output video signal being comprised of a number of pixels that is equal to or less than said first given number but larger than said second given number; and

then further processing said one other output video signal to reduce the number of pixels of the derived image represented thereby by passing a spatially-localized subset of pixels thereof through a spatial window, said subset being comprised of no greater number of pixels than said second given number.

2. The method defined in claim 1, further including the step of:

selectively moving the spatial position of said spatial window within the wide-field-of-view of the derived image represented by said certain output video signal.

3. The method defined in claim 1, wherein:

said first processing step includes the step of deriving from said input video signal an ordinarily-arranged group of a plural number of said other video signals, each other video signal of said group preceding the last other video signal in said group representing a derived wide-field-of-view image exhibiting a higher resolution and a larger number of pixels than the image represented by its next succeeding other video signal in said group, and said last other video signal in said group representing a derived wide-field-of-view image exhibiting a higher resolution and a larger number of pixels than the image represented by said certain video signal; and

said further processing step includes the step of further processing each of said plural number of said other output video signals of said group by passing a spatially-localized subset of pixels of that other output video signal through its own separate spatial window, each of said subsets being comprised of no greater number of pixels than said second given number.

4. The method defined in claim 3 further including the step of:

selectively moving the spatial position of each of said windows.

5. The method defined in claim 4, wherein step of selectively moving comprises the step of:

selectively moving the spatial position of each of said windows independently of one another.

6. The method defined in claim 3, wherein each of said subsets is composed of said second given number of pixels.

7. The method defined in claim 1, wherein:

said first processing step comprises the step of analyzing the spatial-frequency spectrum of said image represented by said input video signal to derive a plural number of separate video signals that represent an ordinally-arranged group of contiguous subspectra bands of said image spatial frequency spectrum, in which said first band exhibits the highest resolution of all bands of said group and each band of said group succeeding said first band exhibits a lower spatial resolution than its immediately preceding band of said group, and said separate video signal that represents the last band of said group constitutes said certain video signal and a given ordinal one of said bands of said group preceding said last band thereof constitutes said one other video signal.

8. The method defined in claim 7, wherein said further processing step comprises the step of further processing each of those separate video signals that represent a band of said group that precedes said last band of said group by passing a spatially-localized subset of pixels of that separate video signal through its own separate spatial window, each of said subsets being comprised of no greater number of pixels than said second given number.

9. The method defined in claim 8 further including the step of:

selectively moving the spatial position of each of said windows.

10. The method defined in claim 9, wherein step of selectively moving comprises the step of:

selectively moving the spatial position of each of said windows independently of one another.

11. The method defined in claim 10, wherein each of said subsets is composed of said second given number of pixels.

12. The method defined in claim 1, wherein said first processing step includes the steps of:

detecting features-of-interest contained in the image represented by said input video signal thereby to derive a detected video signal representing a high-resolution feature-of-interest image defining substantially solely the relative spatial positions of the pixels in the image area occupied by those predetermined features of the image represented by the input video signal which are of interest, whereby only those pixels of said detected video signal which define a detected feature-of-interest have significant level values of either positive or negative polarity with respect to a zero-level value;

processing said detected video signal to derive an absolute-level-value video signal in which each pixel thereof has an absolute level value in accordance with a certain transfer function of the significant level value of a corresponding pixel of said detected video signal; and

processing said absolute-level-value video signal to derive respectively said certain output video signal and said one other output video signal.

13. The method defined in claim 12, wherein:

said step of processing said detected video signal includes substituting a zero-level value pixel for each of those pixels of said detected video signal which have an absolute-value below a predetermined non-zero threshold absolute-value level.

14. The method defined in claim 12, wherein said feature-of-interest is comprised of a moving object situated within the spatial-region area of the image represented by said input video signal.

15. The method defined in claim 12, wherein said feature-of-interest is comprised of a predetermined pattern shape object situated within the spatial-region area of the image represented by said input video signal.

16. The method defined in claim 12, wherein said feature-of-interest is comprised of an object having a predetermined spatial orientation situated within the spatial-region area of the image represented by said input video signal.

17. The method defined in claim 12, wherein:

said step of processing said absolute-level-value video signal includes the step of low-pass filtering and decimation to reduce the high resolution and pixel density of said first given number of pixels of the image represented thereby to the low resolution and pixel density of said second given number of pixels of the image represented by said certain output video signal.

18. The method defined in claim 17, wherein:

said step of processing said absolute-level-value video signal includes the step of cascaded low-pass filtering and decimation of said absolute-level-value video signal a given number of times to derive thereby an ordinally-arranged group comprised of a plural number of other output video signals and said certain output video signal, in which (a) said first other output video signal in said group represents an image exhibiting the same high resolution and pixel density of said first given number as the image represented by said absolute-level-value video signal, (b) each other output video signal except said first other output video signal in said group represents an image exhibiting a lower resolution and pixel density than the image represented by its immediately preceding other output video signal in said group, and (c) said last other output video signal in said group represents an image exhibiting a higher resolution and pixel density than the low resolution and pixel density of said second given number of the image represented by said certain output video signal; and

said further processing step includes the step of further processing each of said plural number of said other output video signals of said group by passing a spatially-localized subset of pixels of that other output video signal through its own separate spatial window, each of said subsets being comprised of no greater number of pixels than said second given number.

19. The method defined in claim 18, wherein each of said windows is movable and further including the steps of:

displaying a relatively busy image derived from said input video signal,

employing said certain video output signal and said plural number of other output video signals after said further processing thereof to selectively move said windows to localize the spatial position of a detected feature-of interest in said displayed busy image and then to generate in accordance therewith a feature spatial position indicator symbol video signal that designates said localized spatial position, and

employing said symbol video signal to superimpose a display of said symbol at the localized spatial position occupied by a detected feature-of-interest in the displayed busy image.

20. The method defined in claim 12, wherein each of said windows is movable and further including the steps of:

displaying a relatively busy image derived from said input video signal,

employing said certain video output signal and said one other output video signals after said further processing thereof to selectively move said windows to localize the spatial position of a detected feature-of interest in said displayed busy image and then to generate in accordance therewith a feature spatial position indicator symbol video signal that designates said localized spatial position, and

employing said symbol video signal to superimpose a display of said symbol at the localized spatial position occupied by a detected feature-of-interest in the displayed busy image.

21. A foveated electronic camera for deriving a data-reduced video output, said camera comprising:

a high-resolution, wide field-of-view imager means for deriving in real time an input video signal representing all the pixels of each of successive image frames of the spatial region being viewed by said imager means, each image frame represented by said input video signal being comprised of a first given number of pixels;

first means for processing said input video signal to derive therefrom (1) a certain output video signal representing respective derived relatively low-spatial-resolution image frames each of which has a given field-of-view that corresponds to the field of each successive image frame represented by said input video signal, each image frame represented by said certain video signal being comprised of a second given number of pixels smaller than said first given number, and (2) at least one other output video signal representing respective derived image frames each of which has said given field-of-view, each image frame represented by said one other output video signal exhibiting a resolution equal to or less than each relatively high-spatial-resolution image frame represented by said input video signal but larger than each relatively low-spatial-resolution derived image frame represented by said certain output video signal, and each derived image frame represented by said one other output video signal being comprised of a number of pixels that is equal to or less than said first given number but larger than said second given number; and

second means for further processing said one other output video signal to reduce the number of pixels of each derived image frame represented thereby by passing a spatially-localized subset of pixels thereof through a spatial window, said subset being comprised of no greater number of pixels than said second given number.

22. The camera defined in claim 21, wherein said camera is adapted to be used with an image-processing computer responsive to the video output from said camera for deriving at least one camera-control signal, and

said second means includes window means responsive to said one camera-control signal applied thereto for selectively moving the spatial position of said spatial window within the wide-field-of-view of the derived image represented by said certain output video signal.

23. The camera defined in claim 21 wherein:

said first processing means comprises means for analyzing the spatial-frequency spectrum of said image represented by said input video signal to derive a plural number of separate video signals that represent an ordinally-arranged group of contiguous subspectra bands of said image spatial frequency spectrum, in which said first band exhibits the highest resolution of all bands of said group and each band of said group succeeding said first band exhibits a lower spatial resolution than its immediately preceding band of said group, and said separate video signal that represents the last band of said group constitutes said certain video signal and a given ordinal one of said bands of said group preceding said last band thereof constitutes said one other video signal.

24. The camera defined in claim 23, wherein said second means comprises window means for further processing each of those separate video signals that represent a band of said group that precedes said last band of said group by passing a spatially-localized subset of pixels of that separate video signal through its own separate spatial window, each of said subsets being comprised of no greater number of pixels than said second given number.

25. The camera defined in claim 24, wherein said camera is adapted to be used with an image-processing computer responsive to said video output from said camera for deriving camera-control signals;

and

said window means includes means responsive to at least one of said camera-control signals applied thereto for selectively moving the spatial position of said movable windows within the wide field of view defined by the pixels of said last band.

26. The camera defined in claim 25, wherein:

said camera-control signals include a separate control signal corresponding to the movable window for each one of said individual bands; and

said window means includes means responsive to each of said separate control signals for selectively moving said spatial position of each of said movable windows independently of one another.

27. The camera defined in claim 24, wherein each of said subsets is composed of said second given number of pixels.

28. The camera defined in claim 24, wherein said second means further includes:

memory means for storing both said second given number of pixels forming the subset from each of said movable windows and said second given number of pixels from said last band of said group; and

selector switch means responsive to a camera-control signal applied thereto for selectively forwarding the stored second given nubmer of pixels corresponding to solely one of the subsets or said last band of said group as said video output from said camera.

29. The camera defined in claim 23, wherein said analyzing means is a Burt-Pyramid analyzer.

30. The camera defined n claim 23, wherein said analyzing means is a filter-subtract-decimate (FSD) analyzer.

31. The camera defined in claim 21, wherein said first processing means includes an analog-to-digital converter for representing the level of each pixel of said input video signal as a multibit number.

32. The camera defined in claim 21, wherein said first processing means includes:

a feature-of-interest detector for detecting features-of-interest contained in the image represented by said input video signal thereby to derive a detected video signal representing a high-resolution feature-of-interest image defining substantially solely the relative positions of the pixels in the image area occupied by those predetermined features of the image represented by the input video signal which are of interest, whereby only those pixels of said detected video signal which define a detected feature-of-interest have significant level values of either positive or negative polarity with respect to a zero-level value;

pixel absolute value transfer-function means for processing said detected video signal to derive an absolute-level-value video signal in which each pixel thereof has an absolute level value in accordance with a certain transfer function of the significant level value of a corresponding pixel of said detected video signal; and

a low-pass convolution filter/decimator pyramid means for processing said absolute-level-value video signal to derive respectively a given plurality of output video signals including said certain output video signal and said one other output video signal.

33. The camera defined in claim 32, wherein:

said pixel absolute value transfer-function means includes substituting a zero-level value pixel for each of those pixels of said detected video signal which have an absolute-value level below a predetermined non-zero threshold absolute-value level.

34. The camera defined in claim 32, wherein said feature-of-interest detector is comprised of a moving-object detector for detecting a moving object situated within the spatial-region area of the image represented by said input video signal.

35. The camera defined in claim 32, wherein said feature-of-interest detector is comprised of a pattern-shape detector for detecting a predetermined-pattern shape object situated within the spatial-region area of the image represented by said input video signal.

36. The camera defined in claim 32, wherein said feature-of-interest detector is comprised of an oriented filter for detecting an object having a predetermined spatial orientation situated within the spatial-region area of the image represented by said input video signal.

37. The camera defined in claim 32, wherein:

said first processing means includes a video signal source for deriving a digitized video signal input to said feature-of-interest detector, said video signal source including a selector switch for deriving said digitized video signal from said input video signal in one switch position of said selector switch and for recursively deriving said digitized video signal from the video output from said camera in another switch position of said detector switch.

38. The camera defined in claim 32, wherein said camera is adapted to be used with an image-processing computer responsive to the video output from said camera for deriving camera-control signals;

and

said second means comprises window means for further processing each separate other output video signal contained in said given plurality of output video signals from said convolution filter/decimator pyramid means by passing a spatially-localized subset of pixels of that separate other output video signal through its own separate movable spatial window, each of said subsets being comprised of no greater number of pixels then said second given number, and

said window means includes means responsive to each of a separate one of said camera-control signals for selectively moving the spatial position of each of said movable windows within the field of the image represented by said certain video output signal.

39. The camera defined in claim 38, wherein said camera is adapted to be used with display means for displaying a relatively busy image derived from said video input signal;

and

and said computer in response to video output signals passed by said movable windows respectively derives eachof said separate camera-control signals to selectively move said windows to spatially localize the spatial position of a detected feature-of-interest in said displayed busy image and in accordance therewith generating a feature spatial position-indicator symbol superimposed on the localized spatial position occupied by a detected feature-of-interest in the displayed busy image.

40. The camera defined in claim 38, wherein each of said subsets is composed of said second given number of pixels.

41. The camera defined in claim 38, wherein said second means further includes:

memory means for storing both said second given number of pixels forming the subset from each of said movable windows and said second given number of pixels from said last band of said group, and

selector switch means responsive to a camera-control signal applied thereto for selectively forwarding the stored second given number of pixels corresponding to solely one of the subsets or said last band of said group as said video output from said camera.