METHOD OF FORMING AN IMAGE BASED ON A PLURALITY OF IMAGE FRAMES, IMAGE PROCESSING SYSTEM AND DIGITAL CAMERA

• US 20120274814A1
• Filed: 07/12/2012
• Published: 11/01/2012
• Est. Priority Date: 10/12/2005
• Status: Active Grant

• Current Assignee: Active Optics Pty. Limited
• Original Assignee: Active Optics Pty. Limited
• Inventors: WAJS, Andrew Augustine
• Granted Patent: US 8,624,923 B2
• Time in Patent Office: Days
• US Class Current: 348/239
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction

Abstract

Image fusion based on a modified method of frame averaging for noise removal by partly averaging over images having a smaller resolution than the desired resolution of the de-noised image. The set of images which are summed for averaging out noise consists of two subsets. The first set of images has a resolution (in terms of number of pixels) being smaller than the resolution of the images in the second set. The resolution of the images in the second set is the resolution of the ‘high-definition’ de-noised output image. The lower resolution images are up-sampled by scaling their pixel numbers to that desired output image. The gradation of the first set images is also adapted to avoid intensity saturation (flare) due to sommation. Image fusing is also done in fourier space using the high frequency components from the higher resolution images and the lower ones from the lower resolution images.

Claims

1. Method of forming a combined image based on a plurality of image frames, including:

obtaining a first set (18;

30;

47;

64) of at least one array of intensity values for representing an image frame at a resolution corresponding to the number of intensity values in the array, and obtaining a second set (19;

31;

46;

61) of at least one array of intensity values for representing an image frame at a resolution corresponding to the number of intensity values in the array,

wherein the combined image is represented by a final array (20;

42;

54;

71) of intensity values,

wherein at least some of the intensity values in the final array (20;

42;

54;

71) are each obtained by executing a step (29;

41;

57;

70) of summing an intensity value from each of at least one array (27;

39;

56;

69) of intensity values based on at least one array of intensity values in only the first set (18;

30;

47;

64) and an intensity value from each of at least one array (19;

40;

46;

61) of intensity values based on at least one array of intensity values in only the second set (19;

31;

46;

61), wherein, prior to executing the summing step (29;

41;

57;

70), only the intensity values of the arrays in the first set (18;

30;

47;

64) are mapped from a scale within a first range to a scale within a second range.

2. Method according to claim 1, including obtaining a first set (18;

30;

47;

64) and a second set (19;

31;

46;

61) arranged such that the image frames represented by the arrays in the first set (18;

30;

47;

64) are represented at lower resolutions than the image frames represented by the arrays in the second set (19;

31;

46;

61).

3. Method according to claim 2, wherein the at least one array (39;

69) of intensity values based on at least one array of intensity values in only the first set (30;

64) contains coefficients in the spatial frequency domain, wherein the at least one array (4061) of intensity values based on at least one array of intensity values in only the second set (31;

61) contains coefficients in the spatial frequency domain and wherein the intensity values in the final array (42;

71) are formed by coefficients in the spatial frequency domain,

wherein at least one lower order coefficient in the final array (42;

71) is obtained by summing an intensity value from each of the at least one array (39;

69) of intensity values based on at least one array of intensity values in only the first set (30;

64) and an intensity value from each of at least one array (40;

61) of intensity values based on at least one array of intensity values in only the second set (31;

61), wherein at least one higher order coefficient in the final array (42;

71) is obtained on the basis of only arrays (40;

61) of intensity values based on the second set (31;

61).

4. Method according to claim 2, wherein at least some of the arrays of intensity values in the first and second set are obtained by reading out measurement values from an image-capturing device comprising an array of light-sensitive cells, wherein each intensity value in the final array (20;

42;

54;

71) is based on at least one intensity value in an array comprised in the second set.

5. Method according to claim 1, including determining an upper limit of the second range at least partly in dependence on the number of arrays of intensity values in the second set (19;

31;

46;

61).

6. Method according to claim 1, wherein at least one of the arrays of intensity values in the first set (47;

64) is obtained by obtaining a plurality of arrays (45;

60) of intensity values for representing an image frame at a resolution corresponding to the number of intensity values in the array, and by summing an intensity value from each of the plurality of arrays (45;

60) to obtain a corresponding intensity value in the at least one array in the first set (47;

64).

7. Method according to claim 1, wherein at least one of the arrays of intensity values in the first set (18;

30;

47;

64) is obtained by obtaining a plurality of arrays (45;

60) of intensity values for representing an image frame at a resolution corresponding to the number of intensity values in the array, wherein the method further includes summing an intensity value from each of the obtained plurality of arrays (45;

60) to obtain a corresponding intensity value in an intermediate combined array (47;

64), and determining an upper limit of the second range at least partly in dependence on at least one intensity value in the intermediate combined array (47;

64).

8. Method according to claim 1, wherein at least the arrays of intensity values in the first set are obtained by obtaining a plurality of arrays of intensity values for representing colour image frames in a first colour space, and applying a transformation to a plurality of arrays of values in a second colour space, wherein, in the first colour space, an image frame is represented by parameter value combinations, each parameter indicating the intensity of one of a plurality of colour components, whereas, in the second colour space, an image frame is represented by parameter value combinations, one parameter of the combination indicating a hue and at least one of the other parameters being indicative of light intensity.

10. Imaging apparatus, e.g. a digital camera (1), comprising a processor (7,9,11) and at least one storage device (8) for storing a plurality of arrays of intensity values, wherein the imaging apparatus is configured to execute a method according to claim 1.

11. Computer program, including a set of instructions capable, when incorporated in a machine-readable medium, of causing a system (1) having information processing capabilities to perform a method according to claim 1.

9. System for processing arrays of intensity values, each array being suitable for representing an image frame at a resolution corresponding to the number of intensity values in the array,

wherein the system is configured to retrieve a first set (18;

30;

47;

64) of at least one array of intensity values and a second set (19;

31;

46;

61) of at least one array of intensity values, the arrays in the first set (18;

30;

47;

64) and arrays in the second set representing respective image frames, and to form a final array (20;

42;

54;

71) of intensity values representing a combined image, wherein the system is configured to obtain each of at least some of the intensity values in the final array (20;

42;

54;

71) by executing a step of summing an intensity value from each of at least one array (27;

39;

56;

69) of intensity values based on at least one array of intensity values in only the first set (18;

30;

47;

64) and an intensity value from each of at least one array of intensity values based on at least one array (19;

40;

46;

61) of intensity values in only the second set (19;

31;

46;

61), and wherein the system is configured to map, prior to executing the summing step, only the intensity values of the arrays in the first set (18;

30;

47;

64) from a scale within a first range to a scale within a second range.