TWI595444B - Image capturing device, depth information generation method and auto-calibration method thereof - Google Patents

Description

Image capturing device and method for generating depth information thereof and method for automatically correcting

The present invention relates to an image capturing device, and more particularly to an image capturing device and a method for generating depth information and a method for automatically correcting the same.

With the development of science and technology, a variety of intelligent image capture devices, such as tablet computers, personal digital assistants, smart phones, etc., have become indispensable tools for modern people. Among them, the high-end smart image capture device is equipped with a camera lens that is comparable to a traditional consumer camera, and can even be replaced by a few high-end models with near-digit monocular pixels and image quality or 3D images. Features.

In general, when a factory produces a two-lens image capturing device, the corresponding spatial position of the two lenses cannot be accurately set to a preset position. Therefore, during the production process, the factory often tests and calibrates the set dual lens module in advance to obtain a set of factory preset calibration parameters. Then, when the user uses the image capturing device, the image capturing device can use the factory preset calibration parameters to correct the image captured by the dual lens to overcome the lack of precision of the process.

However, the above tests and calibration procedures often cost a lot of production costs. In addition, when the above-mentioned image capturing device is actually used by a general user, the image capturing device often causes displacement or rotation of the double lens due to external factors such as accidental dropping, impact, extrusion, temperature or humidity change. Changes in the position of the space. Once the lens is displaced or rotated, the preset calibration parameters in the factory no longer conform to the current application, and the image capture device cannot obtain the correct depth information. For example, if the horizontal imbalance occurs between the two lenses of the stereoscopic image capturing device, the left and right picture levels that are captured after the imbalance are not matched, which may further result in poor three-dimensional stereo shooting.

In view of the above, the present invention provides an image capturing device, a method for generating depth information thereof, and a method for automatically correcting, which can instantly generate a depth of a shooting scene without pre-factoring adjustment of the image capturing device. Information and the generation of automatically corrected stereo images.

The invention provides a method for generating depth information by an image capturing device, which is suitable for an image capturing device having a first lens and a second lens without prior correction, and includes the following steps. First, the first lens and the second lens are used to capture an image of a scene to generate a first image and a second image of the scene. Detecting a plurality of first feature points in the first image and a plurality of second feature points in the second image to calculate pixel offset information between the first image and the second image, thereby obtaining the first image and The angle of rotation between the second images. The image torsion procedure is performed on the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and the second reference image that are aligned with each other. Calculating depth information of the scene according to the first reference image and the second reference image.

The invention provides a method for automatically correcting an image capturing device, which is suitable for an image capturing device having a first lens and a second lens without prior correction, and includes the following steps. First, the first lens and the second lens are used to capture an image of a scene to generate a first image and a second image of the scene. Detecting a plurality of first feature points in the first image and a plurality of second feature points in the second image to calculate pixel offset information between the first image and the second image, thereby obtaining the first image and The angle of rotation between the second images. The image torsion procedure is performed on the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and the second reference image that are aligned with each other. A stereoscopic image of the scene is generated by using the first reference image and the second reference image.

In an embodiment of the invention, the detecting a plurality of first feature points in the first image and the plurality of second feature points in the second image to calculate a pixel between the first image and the second image The step of offsetting information includes first detecting a plurality of feature points in the first image and the second image, and comparing each feature point in the first image and the second image to obtain a plurality of corresponding feature point combinations, and then obtaining Each of the first feature points and each of the second feature points is respectively a pixel coordinate in the first image and the second image, and the pixel offset information between the first image and the second image is calculated, wherein each corresponding feature point is calculated. The combination includes each of the first feature points and a second feature point corresponding to each of the first feature points.

In an embodiment of the invention, the step of obtaining the rotation angle between the first image and the second image includes: painting each of the first image and the second image according to each of the first feature points and the second feature points. The pixel coordinates and the pixel offset information calculate a rotation angle between the first image and the second image.

In an embodiment of the invention, the image torsion procedure is performed on the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and the second reference image that are aligned with each other. The step includes correcting the pixel coordinates of at least one of the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and the second reference image.

In an embodiment of the invention, the step of calculating depth information of the scene according to the first reference image and the second reference image comprises performing third-dimensional depth estimation by using the first reference image and the second reference image to generate the scenario. Depth information.

In an embodiment of the present invention, when the resolution of the first image is not equal to the resolution of the second image, after the step of respectively generating the first image and the second image of the scene, the method further includes: adjusting the first image. At least one of the resolution and the resolution of the second image is such that the resolution of the first image is the same as the resolution of the second image.

The invention further provides an image capturing device without prior correction, comprising a first lens, a second lens, a storage unit and one or more processing units. The storage unit is coupled to the first lens and the second lens for storing images captured by the first lens and the second lens. The processor is coupled to the first lens, the second lens, and the memory, and includes a plurality of modules, wherein the module includes an image capturing module, a feature point detecting module, an image torsion module, and an image processing module. . The image capturing module is configured to capture an image of a scene by using the first lens and the second lens to generate a first image and a second image of the scene. The feature point detection module is configured to detect a plurality of first feature points in the first image and a plurality of second feature points in the second image to calculate a pixel offset between the first image and the second image Information to obtain a rotation angle between the first image and the second image. The image torsion module is configured to perform an image torsion process on the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and the second reference image that are aligned with each other. The depth calculation module is configured to calculate depth information of the scene according to the first reference image and the second reference image.

In an embodiment of the invention, the image capturing device further includes an image adjusting module. When the resolution of the first image is not equal to the resolution of the second image, the image adjusting module is configured to adjust the first image. At least one of the resolution and the resolution of the second image is such that the resolution of the first image is the same as the resolution of the second image.

In an embodiment of the present invention, the image capturing device further includes a depth calculation module, configured to calculate depth information of the scene according to the first reference image and the second reference image.

In an embodiment of the invention, the first lens and the second lens have different optical characteristics or different resolutions.

In an embodiment of the invention, the first lens and the second lens have the same optical characteristics and the same resolution.

Based on the above, the image capturing device and the method for generating the depth information and the method for automatically correcting the image capture device can use the feature point detection after the image capturing device captures two images by using the dual lens. The pixel offset information and the rotation angle between the two images are obtained, and the two images are aligned to obtain depth information of the shooting scene and generate a stereo image. The image capturing device of the invention can generate the depth information of the shooting scene and generate the automatically corrected stereo image in real time without the need of the factory calibration, thereby saving a large production cost.

The above described features and advantages of the invention will be apparent from the following description.

The components of the present invention will be described in detail in the following description in conjunction with the accompanying drawings. These examples are only a part of the invention and do not disclose all of the embodiments of the invention. Rather, these embodiments are merely examples of devices and methods within the scope of the patent application of the present invention.

1 is a block diagram of an image capture device according to an embodiment of the invention, but is for convenience of description and is not intended to limit the present invention. First, all components and configuration relationships of the image capturing device will be described first in FIG. 1. The detailed functions will be disclosed in conjunction with FIG. The image capturing device of the invention can instantly generate the depth information of the shooting scene and generate the automatically corrected stereo image without the need of adjustment by the module factory, thereby saving a large production cost.

Referring to FIG. 1 , the image capturing device 100 includes a first lens 110 a , a second lens 110 b , a memory 115 , and one or more processors 120 . In this embodiment, the image capturing device 100 is, for example, a digital camera, a monocular camera, a digital camera, or other electronic device having a video capturing function, such as a smart phone, a tablet computer, and a personal digital assistant. limit.

The first lens 110a and the second lens 110b include photosensitive elements for respectively sensing the light intensities entering the first lens 110a and the second lens 110b, thereby respectively generating images. The photosensitive element is, for example, a Charge Coupled Device (CCD), a Complementary Metal-Oxide Semiconductor (CMOS) element, or other components. In the present embodiment, the first lens 110a and the second lens 110b are two lenses having the same resolution and the same optical characteristics. However, in other embodiments, the first lens 110a and the second lens 110b are two lenses having different resolutions or optical characteristics such as different focal lengths, photosensitive dimensions, and degree of deformation. For example, the first lens 110a may be a telephoto lens (Telephoto Lens), and the second lens 110b may be a wide-angle lens (Wide-angle Lens); or the first lens 110a may be a lens with high resolution, and the second lens The lens 110b may be a lens having a low resolution.

The memory 115 is, for example, any type of fixed or removable random access memory (RAM), read-only memory (ROM), flash memory, and hard memory. Disc or other similar device or a combination of these devices. The memory 115 is coupled to the first lens 110a and the second lens 110b for storing images captured by the first lens 110a and the second lens 110b.

The processor 120 can be, for example, a central processing unit (CPU), or other programmable general purpose or special purpose microprocessor (Microprocessor), digital signal processor (DSP), Programmable controllers, Application Specific Integrated Circuits (ASICs), Programmable Logic Devices (PLDs), or other similar devices or combinations of these devices. The processor 120 is coupled to the first lens 110a, the second lens 110b, and the memory 115, and includes, for example, an image capturing module 122, a feature point detecting module 124, an image twisting module 126, and a depth calculating module 128. The depth information is generated according to the image captured by the image capturing device 100. The following is a detailed description of the detailed steps of the method for generating depth information for the image capturing device 100.

2 is a flow chart of a method for image alignment and depth information generation of an image capture device according to an embodiment of the invention, and the method of FIG. 2 may be implemented by various components of the image capture device 100 of FIG.

Referring to FIG. 1 and FIG. 2, first, the image capturing module 122 of the image capturing device 100 captures images of a scene by using the first lens 110a and the second lens 110b to respectively generate the first image of the scene. And a second image (step S202). In detail, the first image and the second image captured by the image capturing module 122 by using the first lens 110a and the second lens 110b are two images captured by the different viewing angles for the same scene, which may be, for example, It is a Live-view Image captured in the preview state. Here, the first lens 110a and the second lens 110b capture images by using the same parameters, for example, and the parameters include a focal length, an aperture, a shutter, a white balance, and the like, which are not limited in this embodiment.

Then, the feature point detection module 124 detects a plurality of first feature points in the first image and a plurality of second feature points in the second image to calculate a pixel between the first image and the second image. The information is offset to obtain a rotation angle between the first image and the second image (step S204), wherein each of the foregoing first feature points has a second feature point corresponding thereto.

In detail, the feature point detection module 124 can use Edge Detection, Corner Detection, Blob Detection or other Feature Detection Algorithm. A plurality of feature points in the first image and the second image are detected. Then, the feature point detection module 124 compares the feature points detected by the first image and the second image, so as to be from the first image and the second image according to the color information of the feature point and the adjacent point. Find multiple sets of corresponding feature point combinations. After comparing the first feature point and the second feature point of each set of corresponding feature points, the feature point detection module 124 obtains the pixel coordinates in the first image and the second image, and calculates The pixel offset information between the first image and the second image. Here, the pixel offset information between the first image and the second image represents the degree of displacement of the first lens 110a and/or the second lens 110b.

Specifically, since the first image and the second image are images captured by the first lens 110a and the second lens 110b having different viewing angles, in an ideal state, the corresponding image in the first image and the second image A feature point and a second feature point will be projected to the same coordinate point under the reference coordinate system after being converted via coordinates. On the other hand, if the corresponding feature points on the first image and the second image are not projected to the same coordinate point under the reference coordinate system, the feature point detection module 124 will obtain the offset of the corresponding feature point combination of each group. The program for image alignment is performed in the subsequent steps.

From another point of view, due to the positional setting of the first lens 110a and the second lens 110b, only horizontal aberration or vertical aberration may exist in the first image and the second image in an ideal state. Assuming that the first lens 110a and the second lens 110b are left and right lenses respectively disposed on the same image capturing plane, the first image and the second image should have only a difference in horizontal position. Therefore, if there is a difference in vertical position between the corresponding feature points on the first image and the second image, the feature point detection module 124 will obtain a vertical offset of each set of corresponding feature point combinations for subsequent steps. The program for image alignment is performed accordingly.

In general, when the lens is displaced, it tends to rotate. Therefore, the feature point detection module 124 obtains the pixel coordinates of the first image and the second image and the pixel offset information between the two, and further calculates the relationship between the first image and the second image. The angle of rotation is used to know the degree of rotation of the first lens 110a and/or the second lens 110b.

Afterwards, the image torsion module 126 performs an image torsion process for the first image and the second image according to the pixel offset information and the rotation angle to respectively generate a first reference image and a second reference image, wherein the first reference image and the first reference image are respectively The second reference images are aligned with each other (step S206). In other words, the image torsion module 126 will correct the image coordinates of the first image and/or the second image according to the pixel offset information and the rotation angle to align the corrected image. That is, the first reference image and the second reference image generated after the correction are projected to the same coordinate point under the reference coordinate system after being converted by coordinates. From another point of view, it is assumed that the first lens 110a and the second lens 110b are left and right lenses respectively disposed on the same image capturing plane, and the first reference image and the second reference image generated after the image is twisted only have horizontal aberrations. .

Next, the depth calculation module 128 performs three-dimensional depth estimation using the first reference image and the second reference image to generate depth information of the foregoing scene (step S208). Specifically, the depth calculation module 128 may perform Stereo Matching on each pixel in the first reference image and the second reference image to obtain depth information corresponding to each pixel. In addition, the depth calculation module 128 can record the depth information in the memory 115 in the form of, for example, a depth map (Depth Map), as an application for image processing.

It is to be noted that, in another embodiment, when the first lens 110a is different from the second lens 110b, the image capturing device 100 further includes an image adjusting module (not shown) for adjusting the first image and Second image. For example, when the resolution of the first image is different from the resolution of the second image, the image adjustment module may capture the first image and the second image in step S202 after the image capturing module 122 captures the first image and the second image. The first image and the second image are adjusted to two images of the same resolution to facilitate detection and calculation of subsequent steps.

FIG. 3 is a block diagram of an image capture device according to another embodiment of the present invention, but is for convenience of description and is not intended to limit the present invention.

Referring to FIG. 3 , the image capturing device 300 includes a first lens 310 a , a second lens 310 b , a memory 315 , and a processor 320 , which are similar to the first lens 310 a , the second lens 310 b , and the memory 315 in FIG. 1 . For detailed description of the processor 320, please refer to the related paragraphs mentioned above, and details are not described herein again. The processor 320 of the image capturing device 300 includes an image capturing module 322, a feature point detecting module 324, an image twisting module 326, and an image processing module 328 for performing images captured by the image capturing device 300. Instant automatic correction. The following is a detailed description of the detailed steps of the method for automatically correcting the image capturing device 300.

4 is a flow chart of a method for automatically correcting an image capturing device according to an embodiment of the present invention, and the method of FIG. 4 may be implemented by various components of the image capturing device 300 of FIG.

First, the image capturing module 322 of the image capturing device 300 captures the image of a scene by using the first lens 310a and the second lens 310b to generate the first image and the second image of the scene respectively (step S402). Then, the feature point detection module 324 detects a plurality of first feature points in the first image and a plurality of second feature points in the second image to calculate a pixel between the first image and the second image. The information is shifted to obtain a rotation angle between the first image and the second image (step S404). Afterwards, the image torsion module 326 performs an image torsion process for the first image and the second image according to the pixel offset information and the rotation angle to respectively generate a first reference image and a second reference image, wherein the first reference image and the first reference image are respectively The second reference images are aligned with each other (step S406). For the processing of steps S402, S404, and S406, refer to the related descriptions of steps S202, S204, and S206 in the foregoing paragraphs, and details are not described herein again.

Next, the image processing module 328 will generate a stereoscopic image of the scene by using the first reference image and the second reference image (step S408). In this embodiment, after the first reference image and the second reference image are aligned with each other, the image processing module 328 may directly output the first reference image and the second reference image as a stereo image. In another embodiment, the image processing module 328 can further adjust parameters such as color, brightness, and the like of the first reference image and/or the second reference image to generate two colors and brightness matching images, thereby generating A natural stereoscopic image.

Similar to the embodiment of FIG. 2, the image capturing device 300 may further include an image adjusting module (not shown), and the function of the image capturing device 300 is the same as that of the image capturing device 300, and details are not described herein.

The foregoing method for generating depth information and the method for automatically correcting can be summarized by using FIG. 5 according to a functional block diagram of an embodiment of the present invention.

First, in the image capturing program 502, the image of one scene is captured by the dual lens to generate the first image A and the second image B, respectively. Then, in the feature point detection program 504, the feature points corresponding to the first image A and the second image B are detected to calculate the pixel offset between the first image A and the second image B. Information and rotation angle, wherein the feature points a1 to a3 correspond to the feature points b1 to b3, respectively. In the image reversal program 506, the first reference image A' and the second reference image B' which are aligned with each other are generated based on the pixel offset information and the rotation angle between the first image A and the second image B, respectively.

In an embodiment, after the first reference image A' and the second reference image B' are generated, the depth calculation program 508 is entered to calculate the depth of the scene according to the first reference image A' and the second reference image B'. Information d.

In another embodiment, after the first reference image A′ and the second reference image B′ are generated, the image processing program 510 is entered to generate the stereo image S by using the first reference image A′ and the second reference image B′. .

In another embodiment, the image processing program 610 can also be followed by the depth calculation program 608, that is, the depth information d can also be used as one of the basis for generating the stereoscopic image S. From another point of view, this embodiment is also an integration of the image capturing device 100 and the image capturing device 300.

In summary, the image capturing device and the method for generating depth information and the method for automatically correcting the same in the present invention can utilize the feature point detection after the image capturing device captures two images by using two lenses. The camera obtains the pixel offset information and the rotation angle between the two images, thereby aligning the two images, thereby obtaining depth information of the shooting scene and generating a stereo image. The image capturing device of the invention can instantly generate the depth information of the shooting scene and generate the automatically corrected stereo image without the need of adjustment by the module factory, thereby saving a large production cost.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and any one of ordinary skill in the art can make some changes and refinements without departing from the spirit and scope of the present invention. The scope of the invention is defined by the scope of the appended claims.

100, 300: image capturing device 110a, 310a: first lens 110b, 310b: second lens 115, 315: memory 120, 320: processor 122, 322: image capturing module 124, 324: feature point detection Measurement module 126, 326: image torsion module 128: depth calculation module 328: image processing module S202 to S208: method for generating depth information, S402 to S408: automatic correction method flow 602 to 610: generating depth information and Automatic correction method flow A: first image B: second image a1 ~ a3, b1 ~ b3: feature point A': first reference image B': second reference image d: depth information S: stereo image

FIG. 1 is a block diagram of an image capture device according to an embodiment of the invention. 2 is a flow chart of a method for generating depth information of an image capturing device according to an embodiment of the invention. FIG. 3 is a block diagram of an image capture device according to another embodiment of the invention. FIG. 4 is a flow chart of a method for automatically correcting an image capturing device according to an embodiment of the invention. FIG. 5 is a functional block diagram of a method for generating depth information and a method for automatically correcting an image capturing device according to an embodiment of the invention.

S202~S208: Method flow for generating depth information

Claims (20)

A method for generating depth information, which is suitable for an image capturing device having a first lens and a second lens without alignment by a module factory, the method comprising the steps of: utilizing the first lens and the second lens An image of a scene is captured to generate a first image and a second image of the scene, and a plurality of first feature points in the first image and a plurality of second feature points in the second image are detected to Calculating pixel offset information between the first image and the second image to obtain a rotation angle between the first image and the second image, wherein the first feature point corresponds to the second feature And performing an image warping procedure on the first image and the second image to generate a first reference image and a second reference image, wherein the first reference is generated according to the pixel offset information and the rotation angle The image and the second reference image are aligned with each other; and the depth information of the scene is calculated according to the first reference image and the second reference image. The method of claim 1, wherein the first feature point in the first image and the second feature point in the second image are detected to calculate the first image and the first The step of the pixel offset information between the two images includes: detecting the first image and the plurality of feature points in the second image; comparing the first image and each of the features in the second image a point to obtain a plurality of corresponding feature point combinations, wherein each of the corresponding feature point combinations includes each of the first feature points and the second feature points corresponding to each of the first feature points; A feature point and each of the second feature points are respectively corresponding to pixel coordinates in the first image and the second image, and the pixel offset information between the first image and the second image is calculated. The method of claim 2, wherein the step of obtaining the rotation angle between the first image and the second image comprises: respectively, according to each of the first feature points and each of the second feature points Calculating the rotation angle between the first image and the second image in the first image and the pixel coordinates in the second image and the pixel offset information. The method of claim 2, wherein the image reversal process is performed on the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and The step of the second reference image includes: correcting the pixel coordinates of at least one of the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image And the second reference image. The method of claim 1, wherein the calculating the depth information of the scene according to the first reference image and the second reference image comprises: using the first reference image and the second reference image A three-dimensional depth estimate is used to generate the depth information for the scene. The method of claim 1, wherein when the resolution of the first image is not equal to the resolution of the second image, after the step of respectively generating the first image and the second image of the scene The method further includes: adjusting at least one of a resolution of the first image and a resolution of the second image such that the resolution of the first image is the same as the resolution of the second image. The method of claim 1, wherein the first lens and the second lens have different optical characteristics or different resolutions. The method of claim 1, wherein the first lens and the second lens have the same optical characteristics and the same resolution. An automatic correction method is applicable to an image capturing device having a first lens and a second lens without pre-modulation by a module factory, the method comprising the following steps: using the first lens and the second lens An image of a scene is captured to generate a first image and a second image of the scene, and a plurality of first feature points in the first image and a plurality of second feature points in the second image are detected to Calculating pixel offset information between the first image and the second image to obtain a rotation angle between the first image and the second image, wherein the first feature point corresponds to the second feature And performing an image torsion process on the first image and the second image to generate a first reference image and a second reference image, wherein the first reference image and the The second reference images are aligned with each other; and the first reference image and the second reference image are used to generate a stereoscopic image of the scene. The method of claim 9, wherein the first feature point in the first image and the second feature point in the second image are detected to calculate the first image and the first The step of the pixel offset information between the two images includes: detecting the first image and the plurality of feature points in the second image; comparing the first image and each of the features in the second image a point to obtain a plurality of corresponding feature point combinations, wherein each of the corresponding feature point combinations includes each of the first feature points and the second feature points corresponding to each of the first feature points; A feature point and each of the second feature points are respectively corresponding to pixel coordinates in the first image and the second image, and the pixel offset information between the first image and the second image is calculated. The method of claim 10, wherein the step of obtaining the rotation angle between the first image and the second image comprises: respectively, according to each of the first feature points and each of the second feature points Calculating the rotation angle between the first image and the second image in the first image and the pixel coordinates in the second image and the pixel offset information. The method of claim 10, wherein the image reversal process is performed on the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image and The step of the second reference image includes: correcting the pixel coordinates of at least one of the first image and the second image according to the pixel offset information and the rotation angle to respectively generate the first reference image And the second reference image. The method of claim 9, wherein when the resolution of the first image is not equal to the resolution of the second image, after the step of respectively generating the first image and the second image of the scene The method further includes: adjusting at least one of a resolution of the first image and a resolution of the second image such that the resolution of the first image is the same as the resolution of the second image. The method of claim 9, wherein the first lens and the second lens have different optical characteristics or different resolutions. The method of claim 9, wherein the first lens and the second lens have the same optical characteristics and the same resolution. An image capturing device that does not need to be calibrated by a module factory, comprising: a first lens; a second lens; a memory for storing the image captured by the first lens and the second lens; and; a processor coupled The first lens, the second lens, and the memory, and including a plurality of modules, the module includes: an image capturing module, wherein the first lens and the second lens capture an image of a scene, The first image and the second image are respectively generated by the feature point detection module, and the plurality of first feature points in the first image and the plurality of second feature points in the second image are detected to Calculating pixel offset information between the first image and the second image to obtain a rotation angle between the first image and the second image, wherein the first feature point corresponds to the second feature The image torsion module performs image reversal processing on the first image and the second image according to the pixel offset information and the rotation angle to generate a a reference image and a second reference image, wherein the first reference image and the second reference image are aligned with each other; and the image processing module generates a stereo image of the scene by using the first reference image and the second reference image . The image capturing device of claim 16, further comprising: an image adjustment module, when the resolution of the first image is not equal to the resolution of the second image, adjusting the resolution of the first image and At least one of the resolutions of the second image is such that the resolution of the first image is the same as the resolution of the second image. The image capturing device of claim 16, further comprising: a depth calculation module, configured to calculate depth information of the scene according to the first reference image and the second reference image. The image capturing device of claim 16, wherein the first lens and the second lens have different optical characteristics or different resolutions. The image capturing device of claim 16, wherein the first lens and the second lens have the same optical characteristics and the same resolution.