Computational imaging system

Abstract

An imaging sub-system, a liquid crystal (LC) element, and a digital focus processor are provided. The LC element is placed in the light path of the imaging sub-system, functioning as the aperture of the imaging sub-system, and includes a periodically patterned electrode which is patterned according to a periodical modulation function and configured to blur an intermediate image captured by the imaging sub-system by applying a controllable voltage thereto. The digital focus processor is configured to deconvolute the periodical modulation function to remove the blur away from the intermediate image and determine an all-in-focus real image.

Description

BACKGROUND

1. Technical Field

The present disclosure relates to imaging systems and, particularly, to a computational imaging system.

2. Description of Related Art

Generally, an image of an object captured by conventional imaging systems is in focus only over a limited object distance range which is known as depth of field (DOF). Therefore, it is difficult to sharply capture object scenes that span large distances. To obtain an extended DOF, one attempt has been made that deliberately blurs an intermediate image captured by an imaging system by placing a coded aperture in the aperture of the imaging system and then digitally removes the blur using reconstruction algorithms. The coded aperture is patterned according to a modulation transfer function (e.g., a delta function). As such, reconstruction algorithms can effectively deconvolute the modulation transfer function and restores the image to a more recognizable likeness of the object with a greater DOF than what that would have been otherwise obtainable. This is known as coded aperture imaging and is one kind of computational imaging system. See Zand, J., “Coded Aperture Imaging in High Energy Astronomy”, NASA Laboratory for High Energy Astrophysics (LHEA) at NASA's GSFC (1996); Levin, A., Fergus, R., Durand, F., Freeman, B., “Image and Depth from a Conventional Camera with a Coded Aperture”, ACM Transactions on Graphics (Proc. SIGGRAPH) (2007); Veeraraghavan, A., Raskar, R., Agrawal, A., Mohan, A., Tumblin, J., “Dappled Photography: Mask Enhanced Cameras for Heterodyned Light Fields and Coded Aperture Refocusing”, ACM Transactions on Graphics (Proc. SIGGRAPH) (2007); and Liang, C. K., Lin, T. H., Wong, B. Y., Liu, C., Chen, H. H., “Programmable Aperture Photography: Multiplexed Light Field Acquisition”, ACM Transactions on Graphics (Proc. SIGGRAPH), Vol. 27, No. 3, Article No. 55 (2008). However, to blur the intermediate image, the coded aperture (e.g., the pattern formed on the coded aperture) also blocks large amounts of light rays incident on the aperture, resulting in large amount of light loss.

Therefore, it is desirable to provide a computational imaging system, which can overcome the abovementioned shortcomings.

BRIEF DESCRIPTION OF THE DRAWINGS

Many aspects of the present computational imaging system should be better understood with reference to the following drawings. The components in the drawings are not necessarily drawn to scale, the emphasis instead being placed upon clearly illustrating the principles of the present computational imaging system. Moreover, in the drawings, like reference numerals designate corresponding parts throughout the several views.

FIG. 1 is a schematic view of a computational imaging system, according to a first exemplary embodiment.

FIG. 2 is a planar view of a liquid crystal (LC) element of the computational imaging system of FIG. 1.

FIG. 3 is a planar view of the LC element, according to a second embodiment.

FIG. 4 is a planar view of the LC element, according to a third embodiment.

FIG. 5 is a planar view of the LC element, according to a fourth embodiment.

DETAILED DESCRIPTION

Embodiments of the present computational imaging system will now be described in detail with reference to the drawings.

Referring to FIGS. 1 and 2, a computational imaging system 100, according to a first embodiment, includes a lens 10, an image sensor 20, an LC element 30, and a digital focus processor 40.

The lens 10 and the image sensor 20 constitute an imaging sub-system. The LC element 30 functions as the aperture of the imaging sub-system constituted by the lens 10 and the image sensor 20 (placed in the light path of the imaging sub-system).

The LC element 30 is a transmissive LC panel that has a periodically patterned electrode 32. The electrode 32 is patterned according to a periodical modulation transfer function (i.e., a spatial function):H(x,y)=cos 2π(s_xx+s_yy),  (1)where an origin of the oxy coordinate system is the center of the LC element 30, the x axis extends along the widthwise direction of the LC element 30, the y axis extends along the lengthwise direction of the LC element 30, s_x is a spatial frequency of the electrode 32 along the x axis, and s_y is a spatial frequency of the electrode 32 along the y axis. Assuming that: (i) the refractive index of the LC element 30 outside the electrode 32 is n₀; and (ii) the refractive index of the LC element 30 at the electrode 32 is n=n₀+Δn, where Δn is the refractive index variance caused by applying a voltage to the electrode 32, the refractive index of the entire LC element 30 can be expressed as a refractive index function:n(x,y)=n₀+Δn×cos 2π(s_xx+s_yy).  (2)

Also referring to FIG. 2, in this embodiment, the electrode 32 is a set of concentric annuluses 322 with uniform distances between each two adjacent annuluses 322. However, the electrode 32 is not limited to this embodiment, but can conform to other configurations, for example, a rectangular spiral line 324 as shown in FIG. 3, a circular dot array 326, or a rectangular block array 328 as shown in FIG. 5.

The digital Focus processor 40 includes a Fourier transforming device 42, a deconvolution device 44, an inverse Fourier transforming device 46, and a refocusing device 48.

The Fourier transforming device 42 is configured for transforming a space domain amplitude function U_I(x,y) of an intermediate image captured by the image sensor 20 into a frequency domain function U_ƒ(x,y), where ƒ_x, ƒ_y are x and y axes variables in the frequency domain, respectively. According to Fourier optics, it can be determined that:

$\begin{array}{cc}{U}_f\left(f_x,f_y\right)=\frac{e^{\left[j\frac{1}{2f}\left(f_x^2+f_y^2\right)\right]}}{\mathrm{j\lambda }f}·\int {\int }_{-\infty }^{\infty }{U}_I\left(x,y\right)e^{-j\frac{2\pi }{\lambda f}\left({\mathrm{xf}}_x+y{f}_y\right)}dxdy,& \left(3\right)\end{array}$ where j is the imaginary unit, λ is a wavelength of light rays that captured by the image sensor 20, ƒ(x,y) is a focal length function of each point (e.g., pixel) (x,y) of the image sensor 20 to bring the corresponding point (x,y) into focus.

In addition, the Fourier transforming device 42 is also used for transforming the spatial function of the electrode 32 H(x,y) into a corresponding frequency domain function: H_ƒ(ƒ_x,ƒ_y).

According to complex optics, the function U_ƒ(ƒ_x,ƒ_y) is the convolution of a function U_S(x,y) and the function H(x,y), that is,U_I(x,y)=U_S(x,y)·H(x,y),  (4)wherein the function U_S(x,y) is a spatial domain amplitude function of a real (final) image of objects. As such, to obtain the real image of the objects, the function U_ƒ(ƒ_x,ƒ_y) must go through deconvolution to obtain the function H_ƒ(ƒ_x,ƒ_y). This is accomplished by the deconvolution device 44. According to mathematics, it can be determined that:U_ƒ(ƒ_x,ƒ_y)=F(U_S(x,y))·H_ƒ(ƒ_x,ƒ_y),  (5)where F(U_S(x,y)) is the Fourier transform of the function U_S(x,y). As such, deconvoluting of the function U_ƒ(ƒ_x,ƒ_y) can be expressed as:F(U_S(x,y))={F}⁻¹(U_ƒ(ƒ_x,ƒ_y))H_ƒ(ƒ_x,ƒ_y).  (6)As such, the blur caused by the electrode 32 is digitally removed.

The inverse Fourier transforming device 46 is configured for inversely transforming the frequency domain function F(U_S(x,y)) into the spatial domain amplitude function U_S(x,y) to restore the real image of the objects.

According to the above, it can be determined that the resulting function U_S(x,y) is a function of three variables: x, y, and ƒ(x,y). Therefore, for each point (x,y) of the real image, the unique in-focus focal length ƒ(x,y) can be determined. The refocusing device 50 is configured to determine the unique in-focus focal length for each point (x,y) of the real image to bring all points of the real image into focus. As such, an all-in-focus real image of the objects can be obtained.

By employing the LC element 30, transmittance of the electrode 32 can be controlled by adjusting the voltage applied thereto. As such, the amount of light loss can be controlled and minimized. Typically, to reduce light loss, a transmittance of the electrode 32 is greater than about 50%.

It will be understood that the above particular embodiments and methods are shown and described by way of illustration only. The principles and the features of the present disclosure may be employed in various and numerous embodiment thereof without departing from the scope of the disclosure as claimed. The above-described embodiments illustrate the scope of the disclosure but do not restrict the scope of the disclosure.

Claims

1. A computational imaging system comprising; an imaging sub-system;a transmissive liquid crystal (LC) element placed in the light path of the imaging sub-system and functioning as the aperture of the imaging sub-system; the LC element comprising a periodically patterned electrode which is patterned according to a periodical modulation function and configured to deliberately blur an intermediate image captured by the imaging sub-system by applying a controllable voltage thereto; anda digital focus processor configured to deconvolute the periodical modulation function to remove the blur away from the intermediate image and determine an all-in-focus real image.

2. The computational imaging system of claim 1, wherein the periodical modulation function is: H(x,y)=cos 2π(sxx+syy)where the origin of the oxy coordinate system is the center of the LC element, the x axis extends along the widthwise direction of the LC element, the y axis extends along the lengthwise direction of the LC element, sx is a spatial frequency of the electrode along the x axis, and sy is a spatial frequency of the electrode along the y axis.

3. The computational imaging system of claim 2, wherein the electrode is selected from the group consisting of: a set of concentric annuluses, a rectangular spiral line, a circular dot array, and a rectangular dot array.

4. The computational imaging system of claim 1, wherein a transmittance of the electrode is controlled above 50% by the controllable voltage.

5. The computational imaging system of claim 1, wherein the digital focus processor comprising: a Fourier transforming device being configured for transforming a spatial amplitude function of the intermediate image and the periodical modulation function into a frequency intermediate image function and a frequency modulation function;a deconvolution device being configured to deconvolute the periodical modulation function using the frequency intermediate image function and the frequency modulation function to produce a frequency real image function;an inverse Fourier transforming device being configured for transforming the frequency real image function into a spatial real image amplitude function; anda refocusing device being configured for determining a focal length of each point of the all-in-focus real image according to the spatial real image amplitude function.