Image capturing device and method for calibrating image deformation thereof

1. A method for calibrating image deformation, for an image capturing device having a first image sensor and a second image sensor, wherein the image capturing device has a current calibration parameter associated with the first image sensor and the second image sensor, and the method for calibrating image deformation comprises:

capturing a plurality of image groups through the first image sensor and the second image sensor, wherein each of the image groups includes a first image and a second image, and the image groups include a reference image group;

detecting whether an image deformation occurs on a first reference image captured by the first image sensor and a second reference image captured by the second image sensor in the reference image group, wherein the current calibration parameter is related to a spatial deposition between the first image sensor and the second image sensor, and the image deformation occurs in response to variation of the spatial deposition between the first image sensor and the second image sensor; and

updating the current calibration parameter according to a plurality of feature point comparison values corresponding to the image groups if it is detected that the image deformation occurs on the reference image group, wherein the current calibration parameter is used for performing an image rectification on each of the first images and each of the corresponding second images,

wherein a step of updating the current calibration parameter according to the feature point comparison values corresponding to the image group further comprises:

classifying each of the feature point comparison values into a plurality of statistics bins according to coordinate positions of a plurality of first feature points of the first images and coordinate positions of a plurality of second feature points of the second images; and

determining whether the feature point comparison values are sufficient for computation according to an amount of the feature point comparison values in each of the statistics bins and a plurality of depth values corresponding to the feature point comparison values, and if yes, updating the current calibration parameter according to the feature point comparison values, wherein the depth values are obtained by performing a three dimensional depth estimation on the first feature points and the corresponding second feature points.

2. The method for calibrating image deformation of claim 1 , wherein the step of detecting whether the image deformation occurs on the reference image group comprises:

detecting a first feature point of the first reference image and a second feature point of the second reference image; and

determining whether an offset between image coordinates of the first feature point of the first reference image and the second feature point of the second reference image exceeds a threshold, and if yes, determining that the image deformation occurs on the reference image group.

3. The method for calibrating image deformation of claim 1 , wherein the step of detecting whether the image deformation occurs on the reference image group comprises:

performing a three dimensional depth estimation according to the first reference image and the second reference image to generate reference depth information of a reference focus target in the reference image group, and obtaining a depth focus position related to the reference focus target according to the reference depth information;

obtaining an autofocus position related to the reference focus target by an autofocus procedure; and

determining whether the depth focus position corresponding to the reference focus target matches the autofocus position, and if no, determining that the image deformation occurs on the reference image group.

4. The method for calibrating image deformation of claim 1 , wherein before the step of updating the current calibration parameter according to the feature point comparison values corresponding to the image groups, the method for calibrating image deformation further comprises:

performing a three dimensional depth estimation on the image groups to generate depth information of each of the image groups; and

determining whether to keep the image group according to the depth information of each of the image groups.

5. The method for calibrating image deformation of claim 1 , wherein the step of updating the current calibration parameter according to the feature point comparison values corresponding to the image group further comprises:

performing a feature point detection on the first images and the second images to obtain the first feature points of the first images and the second feature points of the second images;

comparing the coordinate positions of the first feature points with the coordinate positions of second feature points respectively corresponding to the first feature points to obtain the feature point comparison values of the first feature points and the second feature points; and

recording the feature point comparison values of the first feature points and the second feature points.

6. An image capturing device having a first image sensor and a second image sensor, and the image capturing device comprising:

a storage unit, recording a plurality of modules, and storing a current calibration parameter associated with the first image sensor and the second image sensor; and

a processing unit, coupled to the first image sensor, the second image sensor and the storage unit to access and execute the modules recorded in the storage unit, and the modules comprising:

a capturing module, capturing a plurality of image groups through the first image sensor and the second image sensor, each of the image groups including a first image of the first image sensor and a second image of the second image sensor, and the image groups including a reference image group;

a deformation detection module, detecting whether an image deformation occurs on a first reference image captured by the first image sensor and a second reference image captured by the second image sensor in the reference image group, wherein the current calibration parameter is related to a spatial deposition between the first image sensor and the second image sensor, and the image deformation occurs in response to variation of the spatial deposition between the first image sensor and the second image sensor; and

a parameter updating module, wherein the parameter updating module updates the current calibration parameter according to a plurality of feature point comparison values respectively corresponding to the image groups if it is detected that the image deformation occurs on the reference image group, wherein the current calibration parameter is used for performing an image rectification on each of the first images and each of the corresponding second images,

wherein the parameter updating module classifies each of the feature point comparison values into a plurality of statistics bins according to coordinate positions of a plurality of first feature points of the first images and coordinate positions of a plurality of second feature points of the second images,

wherein the parameter updating module determines whether the feature point comparison values are sufficient for computation according to an amount of the feature point comparison values in each of the statistics bins and a plurality of depth values corresponding to the feature point comparison values, and if yes, the parameter updating module updates the current calibration parameter according to the feature point comparison values, wherein the depth values are obtained by performing a three dimensional depth estimation on the first feature points and the corresponding second feature points.

7. The image capturing device of claim 6 , wherein the deformation detection module detects a first feature point of the first reference image and a second feature point of the second reference image, the deformation detection module determines whether an offset between image coordinates of the first feature point of the first reference image and the second feature point of the second reference image exceeds a threshold, and if yes, the deformation detection module determines that the image deformation occurs on the reference image group.

8. The image capturing device of claim 6 , wherein the deformation detection module performs a three dimensional depth estimation according to the first reference image and the second reference image to generate reference depth information of a reference focus target in the reference image group, and obtains a depth focus position related to the reference focus target according to the reference depth information, the deformation detection module obtains an autofocus position related to the reference focus target by an autofocus procedure, and the deformation detection module determines whether the depth focus position corresponding to the reference focus target matches the autofocus position, and if no, the deformation detection module determines that the image deformation occurs on the reference image group.

9. The image capturing device of claim 6 , wherein the modules further comprise:

a depth selection module, wherein the depth selection module performs a three dimensional depth estimation on the image groups to generate depth information of each of the image groups, and the depth selection module determines whether to keep the image group according to the depth information of each of the image groups.

10. The image capturing device of claim 6 , wherein the parameter updating module performs a feature point detection on the first images and the second images to obtain the first feature points of the first images and the second feature points of the second images, the parameter updating module compares the coordinate positions of the first feature points with the coordinate positions of second feature points respectively corresponding to the first feature points to obtain the feature point comparison values of the first feature points and the second feature points, and the parameter updating module records the feature point comparison values of the first feature points and the second feature points.