The MSC Camera.
Depth of Field Enhancement for the Light Microscope.
A Proposal to Digital Camera OEMs.
2 of 5
1 of 1
Section 1. The Optical Multiplexing Hardware.
(Prime Lens, Beamsplitters, Focus Optics and CCD Arrays).
Image Capture and the Optics of the MSC Camera.
The purpose of the optical hardware is to replicate the three-dimensional image space produced by the prime lens, in order to generate n images representing chosen planes in the object space which are simultaneously and continuously available to the computer/software. One replica of the imaging axis is required for each of the n images.
The use of interference-coated beamsplitters ensures that light lost through absorption is kept to a minimum, enabling close to theoretical beamsplit ratios.
Here is a system which can generate four image axes:
And another which can produce nine images:
Achieving Differential Focus.
A lens separated from a CCD array by a distance equal to its focal length will form an image of very distant objects. As the separation between lens and array is increased, the focused plane in the object space moves closer to the camera. Thus virtually any plane in the object space can be bought to a focus by an appropriate separation of lens and array. Other means of achieving focus may be employed, but for the purposes of the present discussion, fixed optics and a z-translating CCD array are assumed.
In the three axis system below, the multiplexing stage is enabling the production of three identical images of a very distant object:
Focus is achieved on each axis by the addition of a positive lens (or lens system) separated from a CCD array by a distance equal to its focal length. This lens (A) forms the second element of a Galilean telescope system for extending the image-forming rays from the prime lens through the beamsplitters. The negative lens (B) behind the prime lens is the first element.
By increasing the separation of the CCD array from the focus lens by different amounts on axes 1 and 2, three planes in the object space can now be bought to a simultaneous focus.
Images of distant (3), middleground (2) and foreground (1) objects are formed on CCD arrays 3´, 2´ and 1´ respectively, and are simultaneously available to the computer for compositing into a single image, or for the production of quasi-stereoscopic image stacks.
This illustrates an important difference between the way in which depth of focus is achieved with the MSC camera and with a normal camera at reduced aperture. Increased depth of focus in the normal camera is always distributed along the optical axis before and beyond the principal focused point. By contrast, the MSC camera can present to the computer images of objects so widely separated in the object space that no purely optical system could bring them to a simultaneous focus.
The following diagram shows how the MSC camera can also be configured to produce a single zone of extended focus similar to that of a normal camera at reduced aperture.
By setting the lens-to-array distances such that d1>d2>d3, the zones of focus may be made contiguous as shown in the inset graphical representations.
Depth of Field, Resolution and Light Utilization in the MSC Camera.
The geometry of image formation is such that halving the diameter of the aperture of the imaging lens will double the depth of focus -- ie, double the axial distance from the focused plane required for the circle of confusion to reach its nominal value. Halving the diameter of an aperture (equivalent to two photographic stops), also halves the resolution of the image, and reduces its intensity by a factor of four.
Reducing aperture 1:2 (2 stops) increases D o F 2:1, reduces intensity 1:4.
Reducing aperture 1:4 (4 stops) increases D o F 4:1, reduces intensity 1:16.
Reducing aperture 1:8 (6 stops) increases D o F 8:1, reduces intensity 1:64.
Reducing aperture 1:16 (8 stops) increases D o F 16:1, reduces intensity 1:256.
Assuming a diffraction limited lens of f 2.8, the aperture reduction required to produce a four-fold increase in depth of field in a single-axis camera would take the aperture to f 11. This is accompanied by the reduction of image resolution to one quarter, and image intensity to one sixteenth of their values at f 2.8. A point is quickly reached where the depth-enhanced image produced by a single-axis camera is too faint and unsharp to be useful.
A four axis MSC camera using the same lens achieves its fourfold depth of focus increase with each CCD array receiving an image having one quarter of the image-forming light, and with the same high resolution as the prime lens at f 2.8. The result is a composite image in which the focused areas have four times the resolution of the single-axis camera at reduced aperture, and at a light sensitivity enabling imaging at one quarter of the subject illumination required by the single-axis camera.
1. A multi-axis (MSC) camera is the only option if more than a four- to five-fold increase of depth of focus at usable resolution and light level is required. A nine-axis system sees the MSC camera into a region of performance which cannot be matched by any conventional camera.
2. The MSC camera is effectively several cameras in one, sharing a common prime lens, and producing a depth of field enhancement of such a different nature to that produced by a single-axis camera that comparison is made difficult.
3. The MSC camera, given computer power equal to the segmenting/compositing task, produces depth of field enhancement with no loss of resolution, and compared to a single-axis camera, becomes relatively more light-efficient for larger increases in depth of field.