Video camera with electronic picture stabilizer

• US 5,563,652 A
• Filed: 06/27/1994
• Issued: 10/08/1996
• Est. Priority Date: 06/28/1993
• Status: Expired due to Term

• Current Assignee: Sanyo Electric Company Limited (Panasonic Holdings Corporation)
• Original Assignee: Sanyo Electric Company Limited (Panasonic Holdings Corporation)
• Inventors: Kobayashi, Akio, Toba, Akira
• Primary Examiner(s): Garber, Wendy
• Assistant Examiner(s): Ho, Tuan V.
• Application Number:
  US08/266,536
• Time in Patent Office: 834 Days
• Field of Search: 348/207, 348/208, 348/155, 348/699, 348/700, 354/70
• US Class Current: 348/207.99
• CPC Class Codes: H04N 23/68 for stable pick-up of the scene, e.g. compensating for camera body vibrations
• CPC Class Codes: H04N 23/6811 based on the image signal
• CPC Class Codes: H04N 23/683 performed by a processor, e.g. controlling the readout of an image memory
• CPC Class Codes: H04N 5/145 Movement estimation for video coding H04N19/51

Abstract

A video camera including a motion detecting circuit in which correlative values of pixels at a present field are evaluated through the comparison of an image signal of a last field and an image signal of the present field. In the motion detecting circuit, an average correlative value and a minimum correlative value are further evaluated. A microcomputer determines whether or not these correlative values satisfy predetermined conditions for each of detection areas defined within an image field. A detection area which satisfies all the predetermined conditions becomes a valid detection area. A whole motion vector is detected by methods different from each other according to the number of the valid detection areas.

Claims

1. A video camera, comprising:

portion motion vector detecting means for detecting a portion motion vector of each of a plurality of detection areas formed in an image field;

valid detection area determining means for determining whether or not each of the detection areas is a valid detection area through determination whether or not each of the portion motion vectors is effective for detecting an unintentional motion of the camera; and

whole motion vector detecting means for detecting a whole motion vector in accordance with methods different from each other in accordance with the number of the valid detection area determined by said valid detection area determining means,

wherein said whole motion vector detecting means comprises;

first dispersion calculating means for obtaining a first dispersion on the basis of the respective portion motion vectors;

selection means for selecting the arbitrary number of the first dispersions having smaller values; and

first detecting means for detecting the whole motion vector on the basis of the portion motion vectors of the detection areas corresponding to the selected first dispersions.

2. A video camera according to claim 1, wherein said first dispersion calculating means comprises:

first calculating means for evaluating a first absolute value representative of a difference between the portion motion vectors in a horizontal direction of each of the detection areas and a mean value of the portion motion vectors in the horizontal direction of the valid detection areas;

second calculating means for evaluating a second absolute value representative of a difference between the portion motion vectors in a vertical direction of each of the detection areas and a mean value of the portion motion vectors in the vertical direction of the valid detection areas; and

adding means for obtaining said first dispersion by adding the first absolute value and the second absolute value to each other for each of the detection areas.

3. A video camera according to claim 1 or 2, wherein said first detecting means detects the whole motion vector by averaging the portion motion vectors.

4. A video camera according to any one of claims 1, 2 or 3, wherein the number of the detection areas is four, and said selection means selects two first dispersions.

5. A video camera according to claim 1, wherein said whole motion vector detecting means comprises:

first replacement means for replacing the portion motion vector of the invalid detection area with the whole motion vector before the present field or frame when the number of valid detection areas is less than a predetermined threshold value; and

second detection means for detecting the whole motion vector on the basis of the portion motion vector of the invalid detection area being replaced with the whole motion vector and the portion motion vectors of the valid detection areas.

6. A video camera according to claim 5, wherein said first replacement means replaces the portion motion vector of the invalid detection area with the whole motion vector evaluated one field or one frame before.

7. A video camera according to claim 1, wherein said whole motion vector detecting means comprises:

second replacement means for replacing the portion motion vectors of the invalid detection areas with a plurality of whole motion vectors which are different from each other and evaluated before the present field or frame when the number of valid detection areas is less than a predetermined threshold value; and

third detection means for detecting the whole motion vector on the basis of the portion motion vectors of the invalid detection areas being replaced with the whole motion vectors and the portion motion vector of the valid detection area.

8. A video camera according to claim 7, wherein said second replacement means replaces the portion motion vectors of the invalid detection areas with the whole motion vectors evaluated one field or frame before and two fields or frames before.

9. A video camera according to claim 1, wherein said whole motion vector detection means includes fourth detection means for detecting the whole motion vector of the present field or frame by multiplying a predetermined coefficient to the whole motion vector evaluated one field or frame before when the number of valid detection areas is zero.

10. A video camera according to claim 9, wherein said coefficient is a numeral value smaller than 1.

11. A video camera according to claim 10, wherein said coefficient is 0.97.

12. A video camera according to claim 1, wherein said valid detection area determining means comprises:

first means for evaluating correlative values of pixels through comparison of an image signal of the last field or frame and an image signal of the present field or frame;

means for evaluating an average correlative value that is a mean value of the correlative values;

means for evaluating a minimum correlative value that is a minimum value of the correlative values;

means for evaluating a value obtained by dividing the average correlative value by the minimum correlative value;

means for evaluating a gradient associated with the minimum correlative value; and

means for determining whether or not the portion motion vector is correctly detected in accordance with the following conditions;

(A) average correlative value>

first threshold value;

(B) value obtained by dividing average correlative value by minimum correlative value>

second threshold value;

(C) gradient>

third threshold value (when average correlative value≧

fifth threshold value; and

(D) gradient>

fourth threshold value (when average correlative value<

fifth threshold value), said third threshold value being equal to or larger than the fourth threshold value.

13. A video camera according to claim 1, wherein said whole motion vector detecting means comprises:

second dispersion calculating means for obtaining a second dispersion on the basis of the portion motion vectors;

comparing means for comparing the second dispersion with a predetermined threshold value; and

fifth detection means for detecting the whole motion vector by methods different from each other in accordance with a comparison result by the comparing means.

14. A video camera according to claim 13, wherein said second dispersion calculating means comprises:

averaging means for evaluating an average vector that is a mean value of the portion motion vectors; and

calculating means for evaluating the second dispersion by adding absolute values of differences between the portion motion vectors and the average vector to each other, and by dividing an addition result by the number of absolute values.

15. A video camera according to claim 14, wherein said portion motion vector detecting means includes means for detecting X direction portion motion vectors and Y direction portion vectors, wherein said averaging means includes means for evaluating an X direction average vector and a Y direction average vector, respectively, and wherein said calculating means evaluates the second dispersion of X direction component of the portion motion vector and the second dispersion of Y direction component of the portion motion vector.

16. A video camera according to claim 13, wherein said comparison means compares a plurality of threshold values with the second dispersion.

17. A video camera according to claim 13, wherein said fifth detecting means comprises:

coefficient setting means for setting a dispersion coefficient in accordance with a comparison result by the comparing means; and

means for obtaining the whole motion vector by multiplying the dispersion coefficient to a mean value of the portion motion vectors of the valid detection areas.

18. A video camera according to claim 13, wherein said fifth detecting means includes means for regarding a minimum portion motion vector as the whole motion vector when the second dispersion is larger than a predetermined value.

19. The video camera of claim 1, wherein said portion motion vector detecting means includes:

first means for evaluating correlative values of pixels through comparison of an image signal of the last field or frame and an image signal of the present field or frame,

means for evaluating an average correlative value that is a mean value of the correlative values,

means for evaluating a minimum correlative value that is a minimum value of the correlative values,

means for evaluating a value obtained by dividing the average correlative value by the minimum correlative value,

means for evaluating a gradient associated with the minimum correlative value, and

means for determining whether or not the portion motion vector is correctly detected in accordance with the following conditions;

(A) average correlative value>

first threshold value;

(B) value obtained by dividing average correlative value by minimum correlative value>

second threshold value;

(C) gradient>

third threshold value (when average correlative value≧

fifth threshold value; and

(D) gradient>

fourth threshold value (when average correlative value<

fifth threshold value), said third threshold value being equal to or larger than the fourth threshold value.