CA1192774A - Objective refractor for the eye - Google Patents

Abstract

OBJECTIVE REFRACTOR FOR THE EYE

Abstract of the Disclosure An objective refractor for the eye is disclosed in which knife-edge optics are utilized. The knife-edge optics cause characteristic illumination of the retina so that components of sphere and astigmatism can be identified.
Provision for remote reading of the characteristic images is provided with the result that two orthogonally disposed knife-edge images can identify the sphere, cylinder and axis required for prescriptive patterns giving the direction and magnitude of required prescriptive change. A system of at least two orthogonally disposed, (and preferably four), knife edges with weighted lighting is disclosed for detection.
Utilization of the knife-edge images is made possible by the detection of the low light level images at a detector having low noise level. A photo-sensitive element divided into a plurality of photo-discrete segments has light from the images proportionally dispersed over its surface. Such dispersion occurs through a matrix of wedge-shaped segments or alternately in the form of optical elements having cylindrical components. This dispersion of the light when used in combination with push-pull knife-edge patterns herein disclosed produces detectable low level refractive signal.
An embodiment using an optic having a plurality-of side by side optic elements, each element having the effect of crossed cylinders, is disclosed with the detector. Separate independent and non-interactive positional information on one hand, and refractive information on the other hand, is provided. Consequently the disclosed refractor is insensi-tive to adjustment and can accommodate a large range of pupil configuration with insensitivity to local retinal variations in light emission.

Description

~27~L

This application is a division of Canadian Patent Application Serial No. 389,012, filed Oc~ober 29~ 1981.

Thi~ inve~tion relates to objective refractors.
More particularly, this invention discloses an objective refractor utili~ing kni~e-edge op~ics and remote image detection at necessarily low light levels.
Summary of the Pri~r Axt Knife-edge optics have not heretofore been prac-ticall~ used wi~h remote obiectiYe refractors. This is ! because ~he images produced by knife-edge optics in conjunc-tio~ with the eye are of extremely low ligh~ levels. These low light level images are extremely difficulk to remotely detect.
Low light level detectors are subject to ~oise.
~pecifically in detecting acros~ ~ broad detection surface a di~ference o~ photosensiti~ity, ~he impedancP or resistance between ad~acent portions o~ the same photosensitive surface is low. Where the resistance is low, and ~he corresponding ~0 electron v'- -nt high, the ~ignal-to-noise ratio quickly becomes destructive of the image difference trying to be ~ensed. There results a severe practical difficulty in trying to de~ect low light level images.
Objective refractors have heretofore been ~ensitive to ~he positioning of the eye. PrecisP pvsitioning of the eye has been r~guixed before accurate ob; ectiYe refxaction can be made. Automatic positioning has not ~een provided for, especlally in a form where ~he positioning inf~xmation ~2~

is non-interactiv~, separate and distinct from the refractive information.
Moreovert prior art objective refractors have included sensitivi~y to ~he light level r turned from the eye. Where, for example, a retina has a variation across its surface on light returned to the observer, heretofore varia-tions in the prescriptive readings have occurred.
Summary of the Invention An o~jective refractor for the eye is disclosed in which knife-edge optics are utiliæed. The knife-edge optics cause characteristic illumination of the retina so that components of sphere and astigmatism can be identifled.
Provisio~ for remote readi~g of ~he characteri~tic images is provided wi~h the result *hat two or~hogonally disposed knife edge images can identify the sphere, cylinder and axis reguired for prescriptive patterns giving the direction and magnitud~ o required prescriptive change. A system of at least two orthogonally disposedr (and pref~rably four), knife edges with weighted lighting is disclosed for detection.
Utilization of ~he knife-edge images is made possible by the detection of ~he low light level images at a detect4r having low noise level. A photo-sensitive element divided into a plurality of photo-discrete segments has light from the images proportionally dispersed over its surface. Such dispersion orcur~ ~hrouyh a ~atri~ of wedge~shaped se~ments or alternately in ~he ~orm o optical elements having cylin-drical components. This dispersion of the light when used in combination with push-pull knife-edge patterns herein dis-closed produces detectable ~ QW level refractive signal.. An 30 embodiment usi~g an optic having a plurality of side by side optic elements, each element having the effect of crossed cylinders is disclo~ed with the detector. Separate indepen-dent and non-interactlve positional infonnation on onP hand, and refractive information on the other hand is provided.
35 Consequently the disclosed refractor is insensitive to adjust-ment and can a~ommodate a large range of pupil configuration with insensitivity to local retinal vaxiations in light emission .

!, 7~7~
The invention, in accordance with the parent application, may be summarized as providing an apparatus for testing an eye comprising in combination: an illumi-nated light source having a boundary with a knife edge terminator; a view path for viewing an eye immediately over the knife edge terminator; means for projecting the image of the light source proximate the knife edge terminator to an eye for producing in the eye illumination of the retinal plane; projection means for projectiny the observed illumination of the eye along an optical path immediately above the knife edge terminator over the kniFe edge terminator to and on a detector surface; a detector matrix dîvided into four discrete quadrants, each detector matrix quadrant being photosensitive and having its photo~
sensitive ele~ents electrically isolated from the photo-sensitive el~ments of other quadrants- and means -for receiving a signal ~rom at least one electrode connected to at least cne of the quadrants for emitting a signal proportional to the illumination of all the quadrants.
On the othér hand9 the invention of the present application may be summarized as providing an apparatus for detecting low level patterns returning from the eye for test-ing the eye comprising in combination; a detector, the detector inc1uding a plurality of apertures; knife edges aligned to at least some of the boundaries of the apertures and terminating inwa~dly along straigilt lines defining a view path ~o and towards the aperture over the knife edges; at least two of the knife edges facing in opposite directions across the central aperture.

Objects, Features and Advantages It is an object of this invention to disclose a knife-edge test with tell-tale illumination patterns on the xetina of the lluman eye. According to ~his aspect of the invention, a light source with a knife-edge terminator pro-jects collimated rays to the eye. Typically, a pro~ection system is incorporated between the knife edge and the eye and is simultaneously used to project the resultant image from ~he eye to an image detector. The light patterns returned from the pupil of the eye have characteristic shape relative to ~he knife edge. Boundaries between light and dark por-tions of the pupil with components parallel to the knife edge indicate components of sphere and astigmatism. Boundaries with components ~ormal to the knife edge indicate components of astismatism along axes at an angle to the knife edge.
An advantage of utilizing knife-edge testing with respect to the human eye is that a tell-tale pattern of pupil illumination is present, which pattern indicates not only refractive error, but gives ~he sense and magnitude of cor-r~ction required. Conseguently, the output of the detectordoes not re~uire hunting in order to determine optimal correction.
A further object of this invention is to disclose measurement of the human eye ~y objective refraction utili-zing at least a light ~ourcet at least one kni~e edge, com-bined projeetion and reception optics and a photodetector.
The 50UrCe shines into the eye ~hrough an apexture formed such ~hat at least a portion of ~he aperture boundary has a ~traight terminator, ~hereby acting as a knife edge barriex on the outgoing beam. The outgoing beam p~sses through the optics in a projecting cap~city; images on ~he eye and thexe-after is passed to the detector by the same optics acting in a xeception capacity. ~ ~in~le knife edge can be used, and ~unctions as a knife edge for light projected to and return ing fro~ the eye. Indeed any such boundary which is straight and knife edge like in character and which serves as an aperture edge for bo~h outgoing and xeturning light simul t~neously will do, provlding that ~he side of ~he boundary 7s7~

which is clear for the ~utgoing beam is opaque for the return-ing beam and vice versa.
A furthe.- object of this invention i~ to disclose a ~eguence of edge illumi~atio~ of preferably four knife edges for interrogation of the eye. These ~nife edges are prefer-ably divided into opposing pairs. One pair of knife edges is illuminated from opposite directions parallel to a fixst axis; the other pair of knife edges is illuminated from opposite directions parallel to a second axis, thi5 second axis being at right angles to the first axis. This opposing and opposite illumination of knife edges produces a "push-pulll' effect in the resultant images. lmage changes due to changing optical prescription in sphere, cylinder and axis can be segregated out frvm other ima,ge degradations, such as ~5 specular reflection from other portions oi the eye as well as optical flare and the like from within the interrogating optical train. Additionally, reduced sensitivity to eye position is achieved.
An advantage of ~he disclosed push-pull knife edge interrogation of the eye is that two separate and non-interactive information bases are generated. The first is positional information. The second is refractive informa-tion. Each of these respecti~e positional and refractive information ~ases are separate and non-interactive.
A further advantage of ~he disclosed detector is ~hat accurate refractive measurements of t~e eye can ~e taken over a wide area. Th~ ins$rument contains insensitivity to adjustment. Hence, accurate refraction can occur even though relatively substantial movement o~ the patient may take place dur.in~ the measurement.
A urther advantage of the disclosed de ector is ~hat it can accommodate a large range of pupil configur~~
tions~ ~oreover, pupil retinas having irregularities in their light tra~smission to ~he do~mstream detector can be measured. Such refractive measurement is insensitive to local retinal variations in the amount of light returned to the detector.

An advantage of this aspect of the invention is that a single detector can interrogate peripheral illumina-ting ed~es in ~eguence. By th~s sequential i~terrOgatlOn, the components of reguired optical correction can be iden-tified sequentially in magnitude and sense.
An additional advantage is that the knife edges can each be separately provided with frequency coded light.
Simultaneous interrogation of multiple knife edges can occur.
A further object o thi~ in~ention is to disclose a preferred matrix of four knife edges for interrogating the eye. Knife edges are aligned in normally disposed pairs.
An advantage of the disclosed knife edge projection systems and light level detectors is that they can be incor-porated in instruments of varying length. Moreover, and by using infrared illumination~ the subject can -~iew along a first path an illuminated target and be interrogated along ~he same pa~h for perfection of the retinal image. A pre-ferred embodiment of light-emitting diode interrogation in ~he infrared spectrum is disclosed.
~n object of ~his invention is to disclose a pre-ferred detector matrix for detecting low level light return-iny from an eye subject to knife edge testing. According to thi5 aspect of the invention, ~he detector matrix is divided into ~our dis~rete ~uadrantsr Each of these quadrants is photodistinct in that ~he photosensitive elements are elec-trically isolated one from ano~her. By the e~pedient of delivering light to a photodistinct portion, a signal is emitted from the photodetector which has a low signal noise ratio.
A further object of this invention is to disclose in ~ombination with a detector having photodistinct eleme~t~
~pecialized optic~ for the distribution of light. According to ~his aspect of ~his invention, multi-element lens~s are inserted between a low light level image in the pupil of the eye and the detector. When ~he low light level image is ~en~rally located, ligh~ is egually distributed to all four detector ~uadrants. With a linear change of position of the centroid of ~he low level light image, a corresponding linear ~127~7~

change ~f image intensity occurs on all detector quadrants.
The detector emits a signal in proportion to the displaceMent of the centroid of ~he low li~ht level image.
~n advantage of this aspect of the invention is *hat the detector is particularly suited for detecting the center of low light level images such as khose returned from knife edge testing of the eye. Th~ optical center of a low light level image can be rapidly indicated. Corresponding corrections can be applied to ~he eye to determine objec-tively the refractive correction required.
Yet another object of this invention is to disclosea mode of measuring at the detector segments the returned low level light images. Accoxding ko ~his aspect of the i~ven~
tion, a ~I- ; ng process is disclosed in which the imaye on a pair of quadrants is summed and differentiated with respect to the image on a r~mainlng pair of guadrants. By the expe-dient of striking a ratio ~f ~he image intensity differenc2s relative to ~he light received on all guadrants, an image signal is received which is proportional to the displacement of low light level images projected~
Yet another object of this invention is to disclose lens configurations for utilization with low level ~ght detection aspects of this invention. According to a first embodiment, *he resultant knife-edge image is relayed to a matrix of deflecting optical wedges or prisms. This matrix of deflecting prisms varies in deflecting in~ensity as dis-placement is varied from a neutral position.
A further object of this invention is to disclose a class of image disper~ing optics, which optics may be util-ized for the displacement of light with optical detectorspreferably of the discrete photoquadrant variety. According to ~his aspect ~f the inventicn, an optic matrix i5 generated havinq an overall optica~ effect that may best be described using lens optics of ~he cross cylinder variety~ A first group of cylinders ~of ei~her positive or negative power~ are laid in a fir~t direction to in effect generate a first light de~lective effect. A ~econd group of cylinders are laid in another direction (prefer~bly at right angles) and disposed 7~

to generate a second light deflective effect. The cylinders used may be chosen from pairings which are positive and positive, negative and negative, or positive and negative (regardless ~f order3. There results an overall matrix of ~ptical elements, which matrix of optical elements causes distribution of light to each of the quadrants of ph~to-discrete detectors.
An advantage of the disclosed lens elements for utilization with photodiscrete detectors is that the gr~ater the number of discrete elements, the less critical the align-ment of the lens elements with respect to a knife edge becomes. For example, where a large number of randomly placed elements is us~d, the need for precise alignment of knife edges wi~h respect to the elements disappears alto~ether.
Yet another object of ~his invention is to disclose other configurations of lens elements that will serve to distribute light among photodiscrete detector segments în proportion to the displacement of low intensity images. By way of example, conical and randomly aligned prismatic seg-rnents all ~ave an effect which can b~ used with the photo-discrete detectors herein disclosed.
~ n additional and preferred embodiment of ~hi5 invention includes a matrix generated by cylindrical lenses of positive and negative power. These cylinders are laid in side-by-side disposition. Along one side of the lens posi-tive and negative cylinders are aligned in a side-by-side array. Along ~he opposite side of the lPns positive and negative cylin~ers are aligned in a side by-side array at preferred right angles to the first array. There results a matrix of crossed cylinder lenses, includin~ positive sphere, negative ~phere, cylinder in ~ first orientation and cylinder in a second and 90 rotated direction. This specialized lens has ~he advantage of dispersing light evenly in a pattern not unlike that generated by the trace of various ~issajous figures.
~ n ~dvantage of thi5 lens is that when it is com-bined with a ~nife edge cutti~g ~cross the lens matrix, the 7~

knife edge at the boundary can genera~e symmetric patterns for detection. These patterns evenly distribute light over a given area, which distributed light may then be detected to photodiscrete detecting elements.
An advantage of the knife edges utilized with the matrix of cylindrical lenses is ~hat the electxical signal out from the detector is directly proportional ~o the inten-sity of the image and the image displacement. Moreover, extremely low light le~els can be sensed. Segments of the photosensitive surface can all be electrically isol~ted one from another.
An advantage of the cylindrical embodiment is thak the overall projection system reguired for the detection of light is ~hortened. Conse~uently, this projection system lends itself to compaztness in the disclosed detector.
A further object of this invention is to dlsclose a preferred embodiment of the lens elements in ~ront of a four quadrant detector. AGCOrding to this aspect of the inven-tion, negative lens surfaces are distributed in side-by-side random relationship over an optical surface, preferably a refractive surface. Specifically, these surfaces are of random alignment and closely spaced. An easily cons-tructed lens element results.
An advantage of this aspect of the invention is that the optical surface can be easily constructed. For example, it has been found that by utilizing a positive mold, such as a ballbearing impressed upon an optical surface or replicati~g media for an optical surface, one obtains a perfectly satisfactory optical element.
A further advantage of this invention is that the disclosed randsmly made optical surface or "pebble plate"
does away wi*h ~he need for precisely aligning the knife edge with respect to an axis of ~he plate. Instead, both the pe~ble plate and the optic elements utilized ~i~h it can be randomly placed one with respect to ano~her.
A further object of this invention is to disclose a pxeferred embodiment of ~he matrix of cylindrical lenses in combination wi~h a ~nife edge. Light from the Xnife edge is Z~

projected ~hrough the specialized op-tics to the eye and light received from the eye passes again through adjacent portions ~f the speciali~ed cylindrical lens. There results in the passage of light to the eye a Lissajous-like dispersement of light along the knife edge. Consequently, only a portion of the light so projected can be seen over the knife edge. The remaining portions of the light projected to the eye from the knife edge are not returnable to the detectors as the physics of the knife e~ge test renders these rays not visible. The portion ~een over the ~nife edge images back to a position immediately abovP the segment of the cylindrical matrix from which projection originally occurred. At this ~egment of the lens a complimentary deflection of the light occurs. There results an enhanced displacement of the light.
An advantage of this aspect of the invention is ~hat the physics of a knife-edge test is used in combination with the predictable dispersion of light at the knife edge to ~creen out all that light, save and except that which has a desired projection angle which can be seen upon return.
There results a low level light signal of enhanced sensitiv-ity returning from the eye.
A further advantage of this invention is that the returning light hits a segment of the cylindrical matrix lenses, which segment pr~duces a complimentary defection.
This complimentary deflection not only further deflects the light, but prQduces an image center of gravity which is an enhanced, and improved si~nal.
A further object ~f ~his invention is.to disclose a flare control illumination pattern. According to this aspect of the invention, the projected light is weighted in inten-~ity about the center of ~he detector. Preferably, two light 60urces are projected on opposite ~ides of the knife edges being utilized. One area is remote from the ~nie edye, the other area is adjacent the knife edge. Specularly reflected images are a functio~ of ~he illumination of bo~h areas and are symmetric~l or cancelling i.n ~heir effect. These specu lar reflection~ form a uniform background to the detector which can be iynored. The remaining image changes are solely ~z~

a function of the knife edge, which knife-edge images can be utilized to determine the sense of reguired correction.
A further object of ~his invention i~ to disclose a preferred knife edge and aperture combination for a detector utilizin~ the invention set forth herein. According to this aspect of the inven~ion, a detector with five apertures is disclosed. The detector includes a central aperture having a dimension of approximately two units by two units. Four peripheral apertures are placed for the sensing of light with each aperture being on a one by one basis. Knife edges are aligned to each aperture. The central aperture includes four inwardly mounted knife edges about the periphery of ~he two by t~o central aperture. The peripheral one by one apertures in~lude ~aired knife edges. These knife edges are each aligned parallel to a knife edge of the central aperture and faced in an opposite direction.
~ n advantage of ~his aspect of the invention is that all the light sources in the detector head are active.
No light sources are located merely for the emitting of light, which light is not utilized in a knife edge testing.
A further advantage of the preferred detector head is that it is particularly adapted to use in opposing detec-ting configurations. For example, the detector head can be utilized for examina~ion of the produced images on a push pull basis.
A further advantage of the preferred knife edge configuration of this invention is ~hat the eye positional information and ~he eye refractive information are separate and non-interactive.
A fur~her object of this invention i5 to disclose an apparatus and me~hod for lo~ating an eye first for tests.
This apparatus and process utilizes the specialized detector head immediately described above. First, knife edges are illuminated along co~linear borders of the central aperture and ~he two peripheral apertures. The single ~nife edge of the central aperture~ faces in a first direction and is generally of two units of leng~hO The paired knife edges of ~he peripheral aperture face in the opposite dixectlon and are each one unit of length. All knife edges are examined together. The central two unit length of knife edge illumi~
nates ~he eye ~n one ~ide of an ~xis. The paired and per-ipheral portions of the knife edge illuminates the eye on the opposite side of the same axis. Since the eye is illumi-nated from both sides of the optical axes sensitivity to refractive error is eliminated. However, by using parallel spaced apart co-linear borders, both positioning o~ the optical axis to the eye and proper distancing of the eye can occur. There results a detector which is particularly sensi-tive to the placement of the eye in front of it.
An advantage of the disclosed sequence for p9Si-tioning the eye is that prescriptive refractive effects are cancelled. As each of the knie ed~es are opposed and of equal length, the resultant projection of light is not sen-sitive to the particular refractive error possessed by the eye. Instead, the detectors evenly illuminate all classes of eyes and permit these eyes to be centered both transversely and towards and away from the detector.
A further object of this invention is to disclose a particularly suitable knife edge combination, which combina-ti~n is sensitive to prescriptive errors and insensitive to the positioning of ~he eye. Accordin~ to this aspect of ~he invention, portions of the apertures are illuminated at their 25 knife edges. Typically, a ~nife edge faced along the central aperture is illuminated. Corresponding knife edges on the peripheral apertures are illuminated. The corresponding ~nife edges face in the same direction, are parallel, but are separated by the width of the central aperture. There xesults a kniie edge alig~ment all in the same direction.
An advantage of ~his aspect of the invention is that prescriptive refractive effects only are picked up;
effects due to ~he positioning of the eye are in large mea-sure ignored~
Yet a further obje~t of ~his invention is to dis clo~e a ~e~uence of examination of the eye. ~ccording to this aspect of the invention, the eye is first positioned utilizing knife edges illuminated in opposite directions ~4 along co-linear portions of the aperture. Thereafter, k~ife edges aligned in ~he same direction along differing portions of the aperture are illumi~ated. During this last knife edge measurement, the optical prescription of the eye is determined.
An advantage of ~he sequence of examination of the eye using the preferred detector of this invention is that two discrete measurements wi~h the prPferred detector can occur. First, and using knife edge pairs, each member of the pair bei~g co~linear but opposed knife edges, the centroid of the eye is determined. Thereafter, and using different knife edge pairs, each member of the pair being parallel aligned ~paced apart but wi~h knife edges faced in the same direction refractive information i~ determined. This information originates in the difference sens~d at the detector in the light level returned from the eye between the interrogations of`the second and di~ferent knife edge pairs. This differ-ence contains prescriptive information which is insensitive to and separate from the positional information.
A furthex advantage of this invention is that the output of the detector readily adapts itself to driving motors in corrective optics. Motors can be activated to null airs and produce emmetropic refraction of the eye through corrective optics.
An advantage of this apparatus and method is that the eye is first positioned ~ith precision with respect to the ob jective refractor . During this posi tion, ,all ambient optical erroxs in ~he eye are ignored. Thereafter, and once ~he eye i5 properly measured for position, the optical errors of the eye are determined. This is determined even thsugh minute movements of ~he eye being tested may naturally occur.
Such mi~ute movements are ignored.
Other objects, features and advantages of this invention can be under~tood after referring ~o the following specification and attached drawings in which:
Fig~. lA~lH are respective illustrations ~nd pro-jections of light rays t~rough ~he human eye from a ~ni.fe edge and illu~rating in ~chematic form ~he shape of knife~
edge images to be viewed;

27~

Fig. lA illustrates an eye with a "near~
6 ighted" or myopic conditioni Fig. lB is a schematic of ~he characteristic image produced ~y such eye;
Fig. lC is a deflection schematic of a posi-tive spherical lens producing such a condition;
Fig. lD is a schematic of an eye with ~ "far-~i~hted" or hyperopic condition;
Fig. lE is a schematic of the charasteristic image produced by such an eye;
Fig. lF is a vector schematic of a lens for producing such a condition;
Fig. 1~ is a combined vector ~chematic, knife edge and characteristic image schem~tic of an eye having astigmatism oriented along a 45/13~ axes;
and Fig. 1~ is a combined vector schematic, knife edge and characteristic image schematic of an eye having astigmatism oxiented alon~ 0/90 axes;
Fig. 2 is a perspective view of a prior art image detector illustrating an embodiment in which high noise levels are present;
Fig. 3 is an embodiment o~ a low level light d~~
tector according to to this invention wherein an image of a light source is focused to dispersing prism wedges and these wedges proporti~nally displace the resultant image to dis-crete photosensitive surfaces;
Fig. 4A is a perspective view of a sp~cialized cylindrical lens matrix utili~ed with this invention, the cylindrical lens matrix having ~n underlying schematic drawing for explaining the function of the lens;
Fig. 4B is a diagram of illustrated segments of the cylindrical lens, this diagram illustrating respective seg-~ents of positive sphere, negative sphere and two components 5 of astigmati6m along opposite ~es;
FigO S is a perspectiYe illustration of a four element lens projected by a ~pherical l~ns system from a light source to an imaging planei ~2~

Fig. 6 is a perspective similar to Fig. 5 with multiple lens segments being illustratedi Fig. 7 is a perspective view similar to Fig. 6 with three ~nife edges disposed at an angle over the face of the lens element, Fi~s. 8A, 8B and 8C are respective represe~tations of lens elements and resultant images on detecting planes of a plurality of knife edges disposed over the specialized lens element of my invention;
~ig. 9 is a p~rspective view of a low light level detector according to ~he preferred embodiment of this inven-tion, special note being made that ~he resultant matrix of photodiscrete seyments is ~ubject to coordinate transfonma-tion to measure the applicable ~eflection;
~5 Fig. 10~ is a side elevation schematic of a knife edge test on the eye of a myope illustrating the factors involved in the ima~e produced in the eye during ~ife edge testing;
Fig. lOB is an illustration of a knife ed~e with ! 20 the cylindrical matrix of this invention only schematically shown illustrating ~he preferred enhancement of the image utilizing the cylindrical matrix and knife edge in combination;
Fig. 11 is a preferred embodiment of the projection ~ystem of thl5 invention utiliziny 2 projection lens, wi*h weighted ill~mination surfaces being present for both control of 1are and background specular reflection; and, Fig. 12 is an alte~nate embodiment of the system of this invention utilizing a lens matrix to both project light to the eye and receive light ~rom the eye.
Fig. 14A i~ an optical schematic illustrating with respect to the lens element originally illustrated in Fig. 4A
how adja~ent optical elements detour light to particular detector quadrants;
Fig. 14B is an illustration of detector quadrants fabricated from egual cross cylirlders, here sho~n as negative cylinders combining to ~e negative lenses, which detector guadrants in turn may bP divided into four portions wi~h each 7~L

portion detouring -~he light impinging thereon to a particular and discrete detector segment;
Fi~. 14C iG an illustration demonstrating how a multiplicity of elements reduces the cri~icality of k~ife edge alignment with respect to the lens segments;
FigO 15A is a schematic illustration of knife edyes cutting ~he lens element of Fig. 14B with distribution of the light being shown o~er the detector segments;
Fig. 15B is a schematic illustration of displace-1~ ment in the X direction of the image shown in Fig. 15A, andparticularly useful for explaininq ~he weighting of the image with respect to the Figure;
Fig. 15C is ~n illustration similar to Fig. 15B
wi~h ~he displacement of the image here occurring in the Y
direction;
Fig. 16A is a schematic of the improved detector head of this invention illustrating the two by two central aperture, and the four one by one peripheral apertures with the respectiYe alignment of the knife edges set forth;
Fig. 16B is a plan view of the detector of Fig. 16A
illustratin~ the apextures and ~nife edges;
Fig. 16C is an illustration omitting a portion of the optical train and illustrating how the detector of this invention is utilized to place an eye in pxoper position for measurement, ~hree d~tector states being illustrated, the detector states being the eye too close for examination, the eye too far away for e~ination, and the eye properly posi-tioned for examination;
Fi~. 16D is an illustration ~imilar to Fig. 16C
with the knife edges ~eing illuminated in an interrogating 6eguence designed or deteL i n; ng the refractive corrections ~ecessary for $he eye;
Fig. 16F is a perspective embodiment of an eye having imaged light sources therein with ~he light ~ources relayed to a position in ~ront of the specialized optics wi~h resultant projection to a detector illustrated;
Fig. 16F i~ an illuskration of the detector plane illustratiny how ~pecular reflection is eliminated as a consideration where in-terrogation by the objective refractor occurs;
Fig. 16G is a perspective representation similar to Fig. 16E utilizing one knife edge, which knife e~ge when incorrectly placed towards and away from the detector screen produces error in the resultant signal;
Fig. 16H is a view of the detector o:E Fig. 16G;
Fig~ 16J is a perspective view S;m; 1 ~r to Fig. 16E, 16G with the utiliæation of three knife edges being illustrated;
Fig. 16K is a view of the detector surface of Fig. 16J
illustratins the detector correctly placed and focused;
Fign 16L is a view o~ the detector of Fig. 16J showing a placement of the detector in an incorrect alignment with the respective images on the -detector still regis-tering the correct optical prescription;
FigO 17A is a perspective view of the preferred "pebble plate" of this invention wherein side by side negative lens surfaces are impressed on a refractive element; and Fig. 17B is a section through the l'pebble plate'l ~aken along the lines 17B-17B of Fig. 17A, Figs. 18A-18D are respective schematic illustrations of a knife edge and detector sur~ace illustrating the so-called "push pull" knife edge interrogation of the eyeO
Referring to FigO lA~ a human eye E having a cornea C
and a lens L is shown viewing a knife edge K. Knife edge K
includes an illuminated portion 14, an edge portion 15 and a point 16 Ishown by an X) immediately above edge 15 from which observation of the illuminated portion o~ the pupil o~ the eye ~16-~%~
is made. The kni~e edge is typically placed at an opticall~
infinite distance from the eye by the expedient o~ collimating optics (not shown). Alternately, projection of the kni~e edge may occur.to any known optical distance~
It will be appreciated that although the side 14 of knife edge K is illuminated or luminous, this ill~nination terminates along edge 15~ Thus no light can be incident through lens L onto the rear retin~ R of the eye from points above edge 15.

.

-16~-~ ereinafter, when the term "knife edge" is util-ized, it will be understood that three discrete functions are ~eferred to.
First, there is a light source. Secondly, the light source terminates along a boundary defining a straight line or ~nife edge terminator. Thirdly, the knife edge kerminator defines immediately thereover an optical path to a detector element.
The illuminated surface below knife edge 15 will produce illumination on the retina R. Fig. lA assumes that eye E is afflicted with myopia. The image plane 18 of knife edge K through lens L will be in front of the plane of the retina of ~he ~ye. A point along this image will form an illuminated oval shape 20 on the retinal surface of ~he eye.
lS Placin~ an observer at point 16 and having th~
observer peer just over the top of the knife edge, will cause light to be collected from an oval area 21 on the retina of ~he eye.
It will be seen ~hat the area of illumination 20 ~nd the area 21 overlap. This area of overlap is identified by ~he num~ral 24. Rays from area 24 may be traced back to the portion of the lens L that will appear to an observer at 16 to be illuminated. Speci~ically, the light will appear to be apparently from the bottom of lens L.
Referring to Fig~ lB, an image of how 7ens L will appear is drawn. This image of lens L shows the illuminated portion caused by light returning from sector 24 with1n the circle of possible returning light 20 rom point 16 above ~nife edge 15.
I* is important ~o note that ~his view is a char-acteristic of the knife edge. It indicates that lens L is excessively positive and the eye ~: has myopia.
Immediately above Fig. lB is a sch~matic diagram le. Schematic diagram lC illustxates in vector forma~ the excessive positive power of lens Le and/or C in Fig. lA.
Turning to Figs. lD, lE and lF, farsightedness or hypermetropia is illustrated. ~Cnife ed~e K with illluninated portion 14 ~toppiny at terminator 15 projects light to the o~

retina R of an eye through a cornea C and a lens Le. As previously shown, the focal plane 18' is here behind the retina R. Projection of the knife edge to optical infinity is assumed and not shown.
Taking projected light from the eye, ~n oval of illumination 23 from one point of source area 14 will be shown on the retina.
Viewing from a point 16 above the terminator 15 of knife edge K, will allow the person to colleck light from oval area 25. The viewer will see light xeturning from an illuminated portion 23 of ar~a 25.
Fig. lE is a view of lens L and how lens L appears to be apparen~ly illuminated. Referring next to Fig. lF, a s~hematic represenkation of the negative defleckion of the le~s Le or C is illu~trated in vector format.
Referring to Fig. lG, only a schematic representa-tion of a lens L, a knife edge K and a retina R is illus-trated. Lens L is illustrated in the schematic vector format similar to Figs. lC and lF. In Fig. lG, lens L is a cross-cylinder lens having power obliguely aligned to edge 15.This lens has astigmatism along 45~-135~ meridians. Lens L
has a positive power along meridian 30 and a negativ~ power along meridian 31. It will be noted -~hat the respective meridians 30 and 31 are at preferred 45 a~gles to edge 15 of ~nife edge K. Noting ~he meridi~ns 30, 31, ~he deflecting power in the vicinity of ~hese meridians ~an be shown. ~or example, and commencing clockwise from the right, at the ~hree o'clock position 32, light will be deflec~ed down-wardly. At ~he 5iX o'clock position 33, the light will be deflected to the right. At ~he nine o~clock position 34, light will be deflect~d upwardly. Finally, at the 12 o'clock po~ition 35, light will be deflected to ~he left.
~ nalyzing the action ~f such a lens in conjunction wi~h a knife edge K can be quickly understood. Ligh~ on one lateral half of the lens passing above ~he knife edge K will be deflected to ~he e~amined eye where it can be viewed.
Light on the opposite ~egment ~f the lens L will be deflected into the ~nife edge K where it may not be viewed. Conse-z~

guently, the imaqe of the retina R will have a terminator Tat right angles to the edge 15 of knife edge K. One segmen~
of the lens L will be illuminated. The illuminat~:d portion of the lens L is shown at 36. As previously set forth, the terminator will not be sharp but rather have a blurred edge.
The term "terminator" should be understood in this manner as it is used hereafter.
The case of a lens L having 0~90 astigmatism can be understood with reference to Fig. lH. Specifically, in Fig. lH, positive cylinder is placed along meridian 40 which is normal to ed~e 15 of knife edge Ko Negative cylinder is placed along meridian 41 which is parallel to edge 15 of knife edge R. The image at the retina R includes an illumi-nated portion 46 with a terminator T that is parallel to knife edge K.
Referring back to Fiys. lB and lE, it can be seen that the terminators T are in substantially the same hori-zontal direction as the knife ed~e. This being the case, it will immediately be realized that astigmatism with axes either parallel to or normal to the edge 15 of knife edge K
will appear the same as spherical components. Consequently, a~d when utilizing only one ~nife edge, only one component of astigmatism can be measured. The measurements of components of astigmatism normal to or parallel to the knife edge cannot be made. We can only say that the information produced from such a measurement is an indication of a "meridiodinal"
power. This measurement can be shown to make sense and be collated to knife edges ~ having ali~nments nor~al to the edge 15. For example, ~he reader is invited to review my U~S. Patent NoO 4,070,115, issued Ja~uary 24, 1978, wherein ~nife edges of differin~ angles are utilized for ~he testing of common lenses.
Having set forth ~he characteristic light patterns ~hat may be pxoduc~d on ~he retina of ~he human eye with ~nife~edge testing and direc~ly observed, reference can now be made to ~he problems encountered in using knife-edge images for remote detection.

7~7~

Speclfically, and where any kind of an image is projected onto ~he retina of the human eye, the intensity of t~at image must necessarily be low. Where the ima~e is in the visible spectrum, the glare problems on the retina are obvious. Where ~he image is either visible or infrared, the images must be of a sufficiently low intensity so that the eye is not burned. Remembering that the rays are in effect focused by the lens L on ~he retina R of the eye, one can immediately understand that the projected light must simply be of a low light level~
When the optics of the eye are utilized to view the illuminated retina, as in the classical case of conventional objective refraction, only a faint image will be visible.
This faint image must be remotely detected if an objective ref~a~tor is to be automated. Moreover, the edge or "termi-natorl' of the image will be far ~rom sharp. The overall image must then be located on "weighted" basis. The problems associated with the projection of such faint images will now be discussed.
Referring to the prior art apparatus illustrated in Fig. 2, a low level light detector is illustrated. Light source S movable about ~n XY plane P is imaged through a lens L to a photosensitive surface D. Photosensitive surface D
typically includes a single and continuous photosensitive surface, either of the photoco~ductiYe or photoresistivevariety. Typically, such surfaces have a "commonl' first connection 50 and are monitored by evenly spaced electrodes 51, 52, 53, 54.
Terri n~l s 51 54 are symmetrically spaced about ~he periphery of photosensitive surface D. Each of ~he terminals is typically connected by leads to the input of an amplifier 55. Amplifier 55 is of conve~tional design and amplifies the difference i~ electrical ~ignal to produce an output propor-tional to X and Y at 56.
When the embodiment of Fig. 2 is applied to a ~ource S of e~tremely low light level, a difficulty arises.
Typically, all the terminals 51-54 are connected to ~ single continuous and conductive layer of the photosensitive 7~7~

material. All these terminals have substantial conductivity between them. This relatively low resistance and high con-ductivity must be ~ensed at amplifier 55 in order to generate a signal at term;n~l s X and Y which is proportional to the displacement of imaye of source S.
Where a high conductivity and hence low resistance is present across electrical terminals, the interYening random motion of electrons creates noise. This noise when received at amplifier 55 and suitably amplified along wit~
the outputs for X and Y results in a low signal to noise ratio. Signal is rapidly lost as the intensity of source S
~lm;n;shesO For example, where source S images at S' on detector D, ~he predominant signals at terminals 51, 52 could well be lost in ~he resultant noise.
The pxoblem therefore becomes one of designing complimentary optics and photodetectors which suppress the tendency of the detector shown in Fig. 1 to produce resultant noise at low image intensity levels.
I will disclose two embodiments. The firs~ of ~hese embodiments will be illustrated with respect to Fig. 3 and illustrate a first conceived and less preferred way of acquiring low light level sensitivity.
Thereafter, and with respect to the remaining illustrations, I will illustrate a preferred knife edge and lens arrayO This preferred knife edge and lens array illus-trates not only a new and useful lens, but additionally discloses ~he new light detector of my invention.
Referring to Fig. 3, and in understanding my first invention, I will first set ~or~h ~he configuration of a plate W. After discussin~ my plate W, I will ~hereafter set forth the remaining optics and operation of ~he system.
Plate W consists of a matri~ of optical wedges.
This matrix has a first and upper ~ide 60 and a ~econd and lower side 62.
3~ ~sr the convenience o the underst~n~; ng of ~he reader, lens W here is ~hown of composite manufacture~ P~
first x~of prism 64 is p~sitioned in the middle of lens W.

7~

The processing of light received uniformly over the top of prism 64 is easy to understand. A first portion of the light will be directed to detector seyments Dl and D2. A
second portion of ~he light incident upon prism 64 will be deflected to detectors D3, D4.
Turning now to an ou~board prism 65, ik ca~ be seen ~hat this prism 65 only includes one facet. This facet will callse light incident uniformly over the top of prism 65 to be deflected only to se~ments ~1~ D2. No portion of prism 65 is disposed to deflect light to detector segments D3, D4.
Prism 66 on the opposite edge of lens W is config-ured in the opposite direction. Specifically, light passing from the direction of ~ource S through prism 65 will be incident upon detector segments D3, D4; no light will be incident upon detectors Dl, ~2.
The intervening prisms 67 and 68 can now be easily understood. Prism 67 has a first portion biased increasingly in favor of segments D3, D4 and a second portion or slope biased to a lesser extent to deflect light onto -the detector se~ments Dl, D2. Prism strip 68 has segments similarly constrl-cted but biased more in favor of detector segments D3, D4, and less in favor of detector segments Dl, D2 Stopping here and understanding the right hand and upper portion of lens W, it will be immediately seen that the fur~her light is deflected towards the right hand portion of lens W, the more light will impinge on de-tector segments D3, D4 and the less light will impinge on segments Dl, D2.
The intervening prisms 69 and 70 on the opposite edge of lens portion 60 can just as easily be understood.
Prism 69 has ~ first facet biased increasingly in favor of ~egm~nts Dl, D2 and a ~econd facet so biased to a lesser e~tent to deflect light onto detector segments D3, D4. Prism ~tr1p 70 has facets similarly constructed but biased more in favor of detector segments Dl/ D2~ and less in ~avor of ~egments D3, ~.
Stopping here and undexst~n~l ng the left hand and upper portion of lens W, it will be immediately seen ~hat the fur~her light is deflected towards the right hand portion of 7?7~

lens W, the more light will implnge upon detector segments D
and D2 and the less light will impinge upon seg~ents D3, D4.
Segments 62 of ~he lens are constructed in an analogous fashion. ~ere, however, the prisms run left and right. DeElection is divided between detector segments Dl, D~ on one hand and D2, D3 on the o~her hand.
Recognizing that the matrix of prisms is formed by the plate W, it will be seen that each area of the matri~.
consists of the effect of an overlying and underlying prism.
These prisms will deflect light to the dete~tor segments proportional to the location at which a source S is imaged.
Passing onto the remainder of the detector, a source S is schematically shown movable in an XY plane P.
Thi~ source S is imaged through a lens 80 so that the image of ~he s~urce 5 falls upon plate W at S'. Assuming that the image at S' is egual to or larger than one of the areas formed by overlying prisms strips, deflection of the light onto the detector ~egments Dl-D4 will be weighted in accor-dance with the position of the image S' on the plate W. A
20 lens 80' underlies plate W to relay the deflected images to the detector plane. Use of this lens is optional, but not reguired.
Detector D i5 typically a photodetector and can include photoconductive cells, photodiodes, photoresistors, phototransistor~, and any other light sensitive detector.
Specifically, the segments Dl, D2, D3, and D4 are all photo~
discrete; that is to say ~hey are electrically separate one from another~ Each segment Dl-D4 has only one electrical connection and ~he current between "commonl' and the elec-trical connection i6 indicative of ~he amount of light inci-dent upon that particular detector se~ment.
By way of pxeferred example, a photosensitive cell including layers of doped silicon of P and N types bonded to an al~ninum surface with appropriate electrical connectors on top and bottom, ~uch as manufa~tured by the Unit~d Detector Technolo~y Company of Culver City, California can be used.
The amplifier 55 i6 a conventional curren~ to voltaye converter and amplifier.

7~

In operation and assuming that an image S' is projected to lens W, light is proportionately distributed by thP prism segments in the matrix to the xespective detector ~egments Dl-D4. By amplifying and logic circuitry ~kandard in ~he art, a signal indicative of the X,Y, position of the image SI on the lens W is produced. Note that "X" and "Y" as ~hown in Fiy. 3 are along the diagonals relative to the detector boundaries.
It will be noted, that as distinguished from the ~mbodiment of Fig. 2, the respective detectors are photo discrete. The resistance between any two of the terminals is essentially infinite as it constitutes an open circuit. Only the amount of light fallin~ on the detector segments produces the desired proportional current flow. ~ence, and even with incidence of low levels of light, the disclosed detector arrangement is essentially free sf noise from the electrical interaction of ~he detector se~ments.
Turning to Fig. 4A, I will now illustrate the preferred lens array and preferred knife edge. This embodi-ment will first be discussed illustrating the make-up of a new lens utilizing ~i~. 4A. Referring to Fig. 4B, I will illustrate the optical characteristics of each of thD lens segments.
Referring to Fig. 4A, lens V consists of a series of side-by~side cylindxical lens strips. Positive cylin-dric~l lens strips 80 hav~ inserted intermediately negativ~
lens strips 81. These strips 80,81 ~lternate in side-by-side relationship with ~he lens strips ~hemselves e~t~ndin~ along ~he width of the lens parall~l to arrow 86. To~ether the sid ~byDside lenses make up a first half of the lens ~en-erally denominated as 8B.
A ~econd and lower half of ~he lens 89 consists of ~ide-by-side positive lens strips 83 and negativ~ lens strips 84. As wa~ previously ~he case, ~he side-by-side ~trips e~tend across ~he lens parallel to the ~imen~ion arrow 87 and f~rmed toge~her ~he second ~ide of the lens 89.
The reader will realize that the lens here illus-trated has been ~hown of composite make~up. In actual fact, ~2~

the divisions between the cylindrical segments 80, 81 and 83, 84 are not visible. Typically, the entire lens is fabricated from molds and i~ made up of a uniform opti.cal materi~l which can be impressed with the desired shape, such as a lens plastic. As with the earlier example, this optical element may also be fabricated with one flat surface and an opposite composite surface h~ving the desired deflections herein described. Having set forth the make-up of the lens ~ith respect to Fig. 4A, ~he optical effects of the underlyin~
matrix will be set for~h with respect to Fig. 4B.
Referring to Fig. 4B, it will be remembered by those having skill in the optical art that two cylinders of egual powers set at right angles sne to another can combine to be the equivalent of a spherical lens.
Looking at a first segment comprising cylinder segments 80, 83, it will be im~ediately se~n that a positive spherical lens e~fect C~ results from the combination of the crossed cylinders. Conversely, and referrin~ to crossed negative cylindrical lenses 81, 84, it will be just as 20 ~uickly realized that the crossed negative lenses result in a negative spherical lens effect C-.
It will be just as guickly remembered that the combinations of crossed positive and negative cyli~ders have an overall cylindrical effect. In this way, it will be s~en that segments 80 and 84 at ~he juncture where they cross form a combined crossed cylindrical lens Al. Similarly, crossed negative and positive cylinders 81, B3 ~orm a combined cylin-drical lens A2.
Stopping here and referring back to ~ig. 4A, it will be seen ~hat each of the discrete lens segmen~s can now be labeled. They can be labeled according to ~heir power.
As ~he pattern in Fig. 4B is repetitious, such labeling of a 6mall portion of the matri~ continues throuyhout the entire lens.
Returning to Fig. 4B, vari~us parallel ray6 in ~heir passage ~hrough di~crete lens elements have been illus-trated as deflectedO These illustra~ed deflections of light can be u~ed to generate a vectorial description of lens deflection.

Referrlng to the illustrated lens deflections, it will be seen ~ha-t each lens segment shown in Fig. ~B has arrows drawn in ~he corners of a figure, which figure is a projection of ~he area of the segment. These arrows c~n be 6een to be desc.riptive of deflections produced. They will hereafter be used to describe deflection produced by my invention.
Referring to Fig. 5, a point source of liyht S
projects light through a spherical lens L to an image plane D. We all know that for all points within the system, that the light will again project to a center point S' on the image plane D.
We now put in lens element V, which I have invented. When plate or lens V goes in, we have a matrix of four side by ~ide lenses. rnly one such matrix of four lenses is illustrated in Fig. 5. In the preferred embodiment this matrix is repeated many times.
Denominating the respective segments, we can put in the designations C-~, C- for the respective positive and negative spherical lenses. Likewise, we can put in the designa-tions Al and A2 for identifying the astigmatic seg-ments of the lens.
We may study another constraint of ~he system.
Remembering that all points S when imaged through lens L converged on the points S~, we may now ask ourselves what happens to rays passing throu~h neutral points of the lens segments C~, C-, Al and A2. In each case, we find that ~he rays a~ainst must end up on the point S'. '~he question ~hen becomes, how are the r~ ~;n7ng rays deflected~
We know that we can use vector descriptions developed wi~h respect ko Fig. 4 to describe ~he deflection of light. This vector description can be made for each of the lens~s about its neutral point. We ~herefore can seguen-tially describe what occurs at each of the remote ~egments of the C~ lens. Ta~ing the principal ray of the system passing ~hlough point 114, we ~now ~hat in the absence of ~pecialized lens V that impingement would be on point S'. ~owever, and due to the ~ector deflection toward.s ~he center of the spher-ical lens C~, we instead will have incidence upon a point 24.

An analysis of a point diametrically opposite the positive spherical lens C+, can be similarly made. Deflec-tion will occur from the normal impingement S' to a new p~int 25 on ~he image plane.
Slrnilarly, for a point 116 on the plate V, a deflection to the point 26 on image plane S will occur. This deflection detourin~ liyht ~hat was oriyinally intended for point S'. Finally, and from point 117 on lens C~, we find imaginy occurring at a point 27.
We may now discuss the case of a negative lens.
Negative lens C- includes a remote point 115' which point 115' again images at point 25. Similarly, it includes a point 116' and 117' which points again image a~out point S' as previously described.
It will of course be appreciated at this point with respect to the astigmatic segments of the lens Al and A2 that only two remaining deflections may be described. Specifi-cally, these deflections are 115'' nd 11~''' at the respec-tive corners. Light rays at these points will be deflected

2~ to poi~t 25.
It will be hereaftex seen that what results fxom the projection of the source S passing through lens L with the specialized lens V substituted therebetween is an evenly distributed square liyht pattern on ~he focal plane D. This image on the plane D has a square shape. With movements of S
along the X and Y axes, corresponding movement of ~he sguare image on plane P will likewise occur.
Turning to Fig. &, we again have a source S movable in an XY plane. Source S has an image on imaging plane P
thrsugh a len~ L. A specializ~d lens element V causes a de~lection pattern with light contained inside a sguare boundary, as explained in the case of the matri~ of four ~ections.
Lens V i5 divided into lenses C~, C-, Al, and A2 as previously described, thi~ time in a matrix of well over four ~uch ~ectionsO Due to the complexity ~f ~he figure, only ~ome of each of the representative lens segments are labeled with ~he appropriate designations C~, C-, Al and A2.

~ 27~7~

Continuing on with the view of Fig. 6, we note again that all segments of the lens project light in s~lare patterns. The light falls within a boundary of a fiquare delineated by the points 24-27 as previously described.
Similar to the case previously described, we know that where translation occurs, this translation will result in a deflection of the entire square image formed by the boundaries 124-1270 Placement of ~life edges at varying alignments acrDss the lens element can be inst.ructive. Turning to Fig.
7, a source S images through a lens L to an imagins plane P.
Again, the specialized lens V is interposed this lens haYing a configuration the same as previously described in Fig. 6.
This time, however, a ~nife edge is placed acxoss the lens element at p~sition Kl, forming a limiting aperture through which light from source S can pass through lens V and hence be~imaged by lens L on image plane P.
As will hereinafter be more fully set forth, it is required that two conditions be met by a knife edged aperture disposed on ~he lens V.
First, the edge of the aperture must traverse Pqual portions of each o~ the four element types comprising speci-~lized lens V (C+~ C-, Al, A2)o Secondly, the edge of the aperture must be disposed across the lens V, at an especial ~lope to ~he boundaries of ~he lens elements of the matrix and n~t parallel to these boundaries.
A particularly preferred embodiment is a slope of 201. The preferred slope is show~ in Fig. 7. Evexy time the illustrated knife edges traverse two elements disposed in the hoxizontal direction, the knife edges txaverse one element disposed in the vertical direction. Other especial slopes, designed a b, will also obtain ~he desired effect if and only if a is odd, when b is even, or b is odd when a is even, where a and b are whole n~ber~.
Knife edge Kl pas~es ~hrough point 135 on lens Al and point 13G on lens C~ is known from ~he example of Fi~. 5 that at these two points, that it will image at 277~

respective points 1~5, 126 on image plane P. The question ~hen becomes where will imaglng occur medially ~or light rays passing between points 135 and 136, say at point 140. Real-izing that point 140 is the peripheral edge of a negative cylindrical lens C-, the problem is simplified. Specifi~
cally, it can guickly be seen that a full negative deflection will be to the periphery of ~he square at a point 150, Thus, taking the case of parallel rays passing sequentially across a knife edge from the point 135 to the point 136, it will be guickly seen that the light rays will image along a line 125, 150, 126.
Taking the case of knife edge K2 and passing from left to right ~he deflection may be understood by superim-posing thereon a ~imilar ve~torial analysis. Starting at point 141 on the left hand edge of ~nife edge K2, it will be remembered that we are in the middle of a positive spherical segment C~. Deflections will be vectorially distributed towards the neutral portion of the element. Impingement of light at point 151 will result. Taking light incident upon ~nife edge K2 at point 142, it will be seen that this p~int is at the upper segment of a positive spherical lens.
Deflection will therefore be downwardly and to the neutral point of the lens with resultant impingement of the light at a point 152.
2~ At point 143, the light will impinge upon at a boundary between the two lens el~ments, *he boundary here being ~hat of a fully negative lens, C . This fully negative lens will cause light incident at that point to be incident at point 153.
At point 144, it will be noted that knife edge K2 passes through the neutral portion of a negative lens.
Con6equ~ntly and in passing through the neutral portion, it will be incident upon the center of the sguare at the point 5'~ Finally, and in passing point 155, light will be inci-dent on the edge o the sguare at 155. There results the ~hown traced zigzag patte~n of traced R2'.
We now for purposes of instruction ~r~ce the path of ray grazi.ng knife edge K3 as it passes through the ele-~i27~

ment. We note that knife edge K3 begins at point 146. Point146 is a section of a positive spherical lens C~ and projects to point 156 on image plane P.
At point 147 we note that 1:he light ray is at a corner of a positive spherical lens C-~ and a negative spher-ical lens C-. Light projected from point 147 following the same logic as in Fig~ 5 ends up poink 127 on plane P. Light from point 14 plots similarly. This light at a periphery of a negative lens element ends up at point 158. The.reafter, light from p~int 159 deflects to point 159.
We thus have traced knife edges Kl, K2 and K3.
There therefore remains the problem of tracing a more complex array in a ~imilar ~nner~ This has been illustrated with respect to the schematic plots of ~igs. 8A and 8~.
Referxing to Fig. 8A, it is instructive to illus-trate deflections of knife edges disposed along Fig. 8A on ~he sguare image trace of Fig. 8B. Here, the observer will note that the light source S and the lens L have been omitted. All we are now going to view is the knife edge as it is disposed across the lens element V shown on Fig. 8A and the resultant traced pattern as lt appears in Fig. 8B.
Taking a knife edge defined by the points 1~0, 181, 182, 183 and 184, the trace can be rapidly generated. Taking point 180, it is obser~ed that thi5 point is at the edge of a positive spherical lens. Re ~ ~ering tha~ in the absence of plate V it would have been deflected to the center of the diagram at point 195 and reme~ering also that it is given a vectorial deflection by the lens element along ~he diagonal direction, it can be seen immediately that it arrives at point 194. Taking point 181 along the knife edge, 181 will be seen to be a portion at the edye vf a negative cylindrical lens. This point is hori~ontally located from a neutral se~ment of a negative lens C-. Accordingly, ~he lens ray will be incident at a point 191. By the same logic, light rays intenne~iate point 190 and 191 will fall along a ~traight line ::onnecting points 190, 191.
I.ight from poin~ 1~2 will prDject to the upper :righthand corner at p~int 192. Remembering that it w~uld 7~

originally have been directed at point 195 and remembering also that it is at an edge of a lens C~, it will be directed to the upper righthand corner of the diagram.
Light from point 183 wlll be incident upon the same point as light from point 181. Remembering that light at point 183 is on the edge of a positive spherical lens and that the positive sphere is directed to the left, deflec~ion will be to the boundary on the left.
Finally, light from point 184 will project to point 194 which is coincident to previously alloted point 190.
We thus see that light along a knife edge intersec-ting the diagonal points of the lens always plots as a V.
It i~ interesting now to investigate light which passes ~hrough neutral points of the segments of the special-ized iens V. This has been plotted along the line which runs1~6, 188, 185, 189, 187, 188', 189~o First, the case of light at points 185 can be easily demonstrated. In that case, we know that the light will in no way be deflected. No deflection will result at impingement of point 195.
Light incident on the lens of Fig. 8A at point 186 falls on the edge of a positive spherical lens. Falling on ~hat edge, it must be deflected to point 196 on Fig. 8B.
Likewise, light incident at 188 falls on the edge of a negative spherical lens. This neg~tive spherical lens plots out at point 198 on the diagram of 8B. Similarly, light at point 189 falls on ~he opposite edge of a negative lens. This li~ht plots out at point 199 after passing ~hrough the neutral point 195 of the lens. Thus, as the knife edge traverses the negative lens C-, we see that we get a linear deflection ~rom points 198 to 195 and finally to point 199. ~t point 187, we are at the edge of a positive spherical lens. Thi~ will deflect to point 197 as illus-tr~ted in Fig. 8B. Light at p~int 188' will be at the edge of a positive ~pherical lensO This will plot out at point 198'. The traverse of the knife edge from point 188' to point 1~9' mu~t pass through a neutral ~eyment of ~he lens at 1950 It will be found ~hat point 1~8' plots on the lefthand 7~7~

edge of 198' and point 189' plots at the righthand edge at 199'. Thus we see we get a pattern that almost looks like a fi~ur~ ~ drawn with straight lines that repeats upon itself.
It is not unlike a Lissajous pattern drawn with straight lines.
Fig. 8B is written on a background. This back-grsund includes horizontal axes X and vertical axes ~. The figure projects along boundaries 100, 101, 10~, 103 (labeled clockwise).
We can also see that each of the lines traces into respective quadrants of ~hese figures. These ~uadrants ~hemselves can be labeled quadrant 104, 105, 106, 107.
~ n interesting ohservation can be made. The length 4f line xesultant from ~he projections of the knife edge in each of ~he ~uadrants is equal. It is equal in linear length. It is al50 equal in the center of gravity sense.
Specifically, it will be found that the center of gravity of the line segments in all portions of the images falls sym-metrically about point 195.
We no~ go to Fig. 8C. Fig. 8C is a diagram of the matrix of Fig. 8B superimposed upon a detector. The detector includes photodiscrete quadrants Dl, D2, D3 and D4. ~ach of these guadrants has approximately the same area as the boundary square which includes the deflection patterns pro-duced by ~he respective knife edges. At this point, it will be seen that the image in Fig. 8C has been moved along a diagonal 110 to the upper left. As previously illu~trated, the detector segments are photodiscrete or sep~r,ate along lines of division 11~, 115.
In order to measure a deflection of ~he image on a proportionate basis, it is necessary ~hat the amount of line cut from a given ~nife edge always be proportiona~ely distri-buted in each of ~he detector segments Dl-D40 This propor-tionate distribution should be egual to the direction and amount of displacemenk w~ich has occurred. Therefore, where a displacement i~ along and parallel to a diagonal 110, respective detector ~egments Dl and D3 should have e~ual amounts of light incident upon them. The.re should be no 7~

difference in signal registered between them to indicate a displacement other ~han along diagonal 110.
In Fig. 8C, the trace of the knife edge of point 18~, 181, 1~2, 183, 1~4 has heen generated. This trace plots is given the same numeric design~tion.
It can be demonstrated and is indeed apparent from a visual inspection of the drawing, that the linear length of light line appearing in detector segments Dl and D3 is equal.
The linear length of light line appearing in segments D2 and D4 is not equal. The difference is proportional to the displacement as it is occurred along the diagonal 110. Plot of the knife edge designated by points 186, 188, la5, lB9, 187, lB8l, 185, 189' yields ~he same results, and it will be found that the amount of line residing in detector 6egments Dl and D3 is ~h~ same. The amount of light line rernaining in detector segments D2 and D4 however is again different and by the same amount as before.
Displacement along the opposite diagonal 111 will yield a similar result. Moreover, I have found that dis-placements on any direction followed the above rule. Thedifference in the amount of light line that is laid down between any opposite quadrants will be proportionate to the displacement. It i~ this result which allows me to apply khis detectox for the detection of low level light sources with photodis~rete detector segments.
It will be seen ~hat ~he center of gravity 195 or S' will thus be tracked in its displacement according to ~he difference in amount of light received at each of the detector segments. It is there~ore possible to get a linear out~ut.
Putting an infinite number of ~nife edges ox narrow bands of light across ~he lens elements, it will be immedi-ately realized that the result will be a solid, evenly dis~
tributed patch of light inside a boundary of the ~ame shape as ~he lens elements. This patch of light will be the con-jugate i~age of every point 60urce of light in a faint and measured image. By utilizing a summation of these conjugate distributed images, each ~ounded in a square, I have a pecu-2~7~

lia.rly useful detector image which incident upon a detectorplane will read out X and Y positions for the center of gravity of a faint and remo~e image. It is this character-istic of being able to recognize khe center of gra~ity of a faint image that enables this detector to be peculiarly useful.
Raving described ~he construction of the lens element and the deflection that is utilized within the lens element, the apparatus of Fig. 9 can now be set forth.
Referring to Fiy. 9, a light source S is illustrated in XY
plane P. This source S projects past a lens L and lens element V. Lens element V projects an image of light onto a detector ~urface D having photodiscrete guadrants Dl-D4.
In the embodiment of Fig. 9, it will be noticed ~hat source S illuminates ~he upper righthand quadrant of plane XY. ~he low level intensity image is projected from source S through ~he combined lens L and specialized lens V.
Specialized lens V is surrounded by knife edges Kl-X4. These respective knife edges all establish an opague terminator to the othexwise transparent lens V previously described.
Two optical effects are present when source S
projects its li~ht past lens V and the knife edges Kl-K4.
First, ~he knife edges when projected to the sur-face of the detector D including the photodiscrete segmentsDl-D4 are at an angl~ to ~he sguare sides cont~in;ng the illumination.
Secondly, the resultant light from any point on the image forms an evenly distributed sguare image, which ~enly distributed ~quare image is translated on the detector se~-~ents in accordance with ~he translation of the source S at the plane P. Thus, where ~he source S moves to the upper right hand guadrant of ~he source P in Fig. S, ~he sgu re patch of light would move to the lower left relati~e to an XY
plane. ~oving to the lower left relative to an X~ plane, the detector of Fig. 9 when connected to a standard circuit such as ~hat sho~n in the amplifier o Fig. 1 can read out in the XY position.

~27~

It will be realized, however, that due to the properties of the image, a coordinate transform will have to be applied as the edged direc~ionci and coordinate directions will differ. Since such coordinate transfo~ns are well-known in the art, they will not be repeated here.
The disclosed lens element has an unexpected result, when utilized to project light and receive light over a ~nife edye to and from an eye. Fig. lOA is a schematic diagram of light from a lunife edge test impinging upon ~he ey~ of a myope. Fig. lOB is a schematic illustrating the principle of how as light comes to a focus a signal enhancing displacement occurs.
Taking the c~se of the eye previously illustrated in Fig. lA, it will be remembered that this eye suffer2d from 1~ the vision defect of myopia.
Returning to Fig. lOB, a series of light rays passing from knife edge K can in sequence be considered.
Each of these light rays when passing from the knife edge must first pass through lens V. In passing through lens V, the light rays dependent upon their respective left to right points of origin encounter from left to right across the top of the knife edge lens se~nents Al, C~, C- and A2 at the lens V point of meridiance.
Referring t~ Fig. lOA,! a sc~ematic of the knife edge test of Fig. 1 on ~he eye of a myope is illustrated.
This figure illustrates the physics of the resultant rather indefinite image produced on the retina. A knife edge K
ill~inated at a portion 250 below a terminator ~51 i~ imaged ~hrough the lens L of ~he myope. This produces in accordance with the myopic deficiency of the eye E an image of the knife edge K' in fron~ of the retina plane R.
Viewin~ the respective points on which an image of ~he knife edge te_ ;n~tor 251 can be projected through 3 points on the eye can be in~tructive. First, and throuyh the central portion of the eye, 262 it will be seen ~hat the illuminated ~nife edge ~50 will be projected on the retina khrough arl enlarged illurninated are~ 262'. Secondly, ~he ~ame knife edge when projected thr~ugh point 261 on ~he eye 77~

will be projected ~hrough an additional and enlarged area 261'. Finally, projection through point 263 will produce an enlarged image Z6~'. Thus the total image will be spread over an enlarged area of the eye, which area ~f the eye must then, in accordance with the limitations of knife edge imaging, be viewed over the top of the knife edge termina'cor 251. This will be ~he portion immediately over the term-inator 251.
Constructing a straight line from point 261 to and across the image of the knife edge to retina of the eye, one immediately can determine a terminatox of that portion of the retinal plane which may be viewed. Constructing a terminator of the viewed area over ~he knife edge, one can project an image of the terminator at 252~. Constructing kerminators from point 263 ~hrough the terminator image 252' to the retina give~ a window throuyA which light impinging on the retina may be returned immediately over the knife edge K.
- It will be appreciated that the terminator of the image on ~he retina will be indefinite and out of focus. As correction is made to the eyes of the myope through inter-vening optics, the image K' of the knife edge will approach the retina R of the eye. As it approaches the retin~ R of the eye, the terminators will sharpen. When the terminators sharpen, the unexpected result of utilizing the displacing lens to pro~ect light to ~he eye and receive light back from the eye ~ill be ~nh~n~ed wi~h ~he sha~pness of the im~ge ter~i natc: r .
In encountering these respective seyments Al, C~, C- and A2, the light will be deflected as it passes immedi-30 ately over ~he edge of the knife in the patterns previously described with respect to Figs. 8A and 8B. The light will attempt to generate a square patt~rn on the lens L ~f the eye E and finally pass to ~he retina of the eye R where ~he myopic condition is illustrated.
Xnife edge tests even through a specialiæed element ~uch as the element V have one ~hing in common. ~his factor i~ that light returning ~o a ~nife edge always returns to a spot immediately adjacent the light area fr~m which light was ~z~

orlginally emanated assuming a moderate sta~e of refractiveerror. Thus in ~he illustrated case, light emanating from ~he illuminated edge of ~he knife (the reverse edge in the illustration of Fi~. 10B) will return to the knife edge K at a position immediately above C~. The light will pass through ~he particular lens segment Al, C~, C , or A2:
Ohserving fuxther ~he diagram of the myopi.a illus-trated in Fig. 10B, we know that the li.ght incident upon an area 24' will return from an illuminated ar~a 24 from the lens L of the eye E. It will return and again receive an upward deflection. When it receives this upward deflection, it will pass to a detector.
Two effects will occur because of ~he passage of light to lens L of eye E through the specialized lens V.
First, rays deflected by the elements of the lens V
to any portion ~f ~he eye other than the upper portion 24' will never ~e seen. Thus, the total amount of light received back from the ~nife E over ~he top of the knife edge will be ~1 lnlshed; only those rays which are emanated to ~he upper poxtion of the eye will have enhanced reception upon their return .
Secondly, and since in knife edge testing of ~he eye rays return from diametrically opposite portions of the eye, light rays will have a greater total deflection when ~5 received back ~rom the eye~
There results an image of increased deflection with increased contrast.
Another way to understand this aspect of my inven-tion is to analyze the case of parallel rays ~equentially left to right leaYing ~he knife edge. Upon passing through th~ speciali~ed len~ or "wobble plate" V, all ~he parallel rays will be ~prayed in patterns, which patterns have been previously illustrated. Only that portion of ~he pattern which is ~prayed to the upper portion of the eye L will be seen over at the corresponding point along the top 9f the knife edge K upon return. Moreover, ~he portion that is returned will be returned from the lower ~e~ment ~f the eye 24 and have a ~econd deflection upwardly upon passing by the 7~7~

knife edge K for the second time. ~his second deflection when received at a photodetector such as that illustrated in Fig. 11 will give enhanced contrast through enhanced light ray displacement in analyzing the resultant image.
Review of ~he images returned from the eye by other optical defects is analogous. In each case, the light that can be accepted from a knife edge test enters the eye at one portion and exits at a diametrically opposite portion. It can therefore be seen that ~he enhanced deflection principle above-entitled will work for all vision defects. For exam-ple, in the case of "farsightedness" illustrated in Fig. lE, light entering the bottom portion of the lens 23' will exit the top portion 23~ Likewise and with respect to Fig. lG, li~ht entering the lefthand segment of the lens L at 36' will exi~ area 36. The resultant enhAnced deflection will be the same.
Referring to Fig. 11, the specialized lens V of this invention is shown placed over a detector aperture 200.
Aperture 200 is surrounded by four knife edge pairs, the respective kni~e edye pairs being denominated by the desig-nations A, A', B, B', C, C' and D, D'.
Observing ~hese knife edges placed in a square pattern about detector aperture 200, it will be noticed that only the light emitting apertures A, B, C and ~ are immedi-ately adiacent the detector aperture 200. These lightsources having their ~dge adjacent the aperture 200 form the four knife edges previously illustrated.
It has been found in addition to the retinal reflections obser~ed, there will be certain corneal and iri~
reflections going back to ~he detector Dl. If only one side of the detector aperture is ill~ninated, one ~nife edge will have the effect of weighting the image received at ~he detec-tor segments Dl, D2~ D3, D4. Since this is the case, it has been found expedient to illuminate the knife edges in pairs.
l'hus when ~life edge segment A is illuminated, segment A' is also illuminated.
Regarding segment A', it will be noted that it is ~eparated a distance from the ~nife edge formed by light ~2~

element C. Since it is separated by the width of the element C from the detector aperture ~00, substantially no light will return from source A' due to the retinal k~ife edge effect.
The only light that will return will be that light which is from other reflected sources, ~uch as corneal reflections, iris light, and the like. In order to relay light from the knife edges and to the eye, and from the eye to the detector, a lens 203 may be optionally placed between the light sources and eye.
In order to assure that the combinations of illus-trated light sources A, A' contribute no weight to the over-all displacement of ~he image, both light sources are given an effectivity which is s~mmetric to ~he center of the li~ht-recei.ving aperture 201. In order to do this, light source C
is given an intensity slightly greater than light source C';
this intensity is such that the product of the distance from point 201 to light source C eguals to the product of the distance from point 201 to light source C'. Naturally, the same illumination scheme is utilized in lîght sources B, B';
C, C'; and D,D'.
Relay of the image t~ the eye E is shown occurring via a lens 203. This relay system is only schematically illustrated. Any number of relay systems can be used.
It will be observed that each of the light sources 2~ A-D' is covered with a portion of a lens. Preferably, the cylindrical lens is gi~en a focal length so that in combi nation with the other optics, the knife edge is projected to the retina R of the eye E. Light returning from the faint image of ~he retina ~ of the eye E will pass through the lens element V, the detectDr aperture 200 and to and on the detec-tox segments Dl-D4 previously described.
Referring to Fig. 12, a preferred embodiment of my objective refractor is disclosed~ According to ~his embodi-me~t a wobble plate W is illustrated overlying not only the detect~r aperture 200 but additionally each of the light ~ources as well. Re~ultant deflection from each knife edge occurs as it is illustrated schematically with re pect to Fig. 10. Thus, each of khe four knife edges has an optical 7'7~

pattern imaged to the eye and each of the ~ptical edges in return passes light to the detectvr segments Dl-D4 in the ma~ner pxeviously illustrate~ It can be thus seen tha the plate W herein can be oper~ble either over that portion of the knife edge emitting ligh~ to the eye, that portion of the knife edge receiving light froM the eye, or both (as illus-trated in Fig. 12).
During the development of this invention, I have made a surprising discovery. Specifically, I have determined that any optical element composed of cross cylinder lenses is sufficient for ~he practice of this invention. I have fur-~her determined that the cross cylinder lenses can be formed from any repetitive combination of cylinders including the case where the cylinders are positive and positive, negative and positive, positive and negative, and/or negative and negative. Specifically, and wi~h respect to matrices com-posed of negative lenses, I find these to be a preferred embodiment, especially if they are placed in a random pattern with respect to the knife edge.
I have further determined that other optical sur-faces will work for the distribution of light. So long as the light is evenly distributed ~rom a central detector position to all detector quadrants and light is proportion-ally moved between the detector se~ments with detected image movement, an optic element cont~; ni ng multiple deflecting facets will work.
By use of the word optic, I intend to cover both mirrors and lenses. By use of the word de~lection I intend to cover both refraction and reflection.
As an example of the diverse surfaces whi~h may be used, cylinders, randomly aligned pyramids and ~he like may all be utilized as the deflecting ~urfaces.
Referring to FigO 14A, I have caused a diagxam to be ~isplayed illustratiny ne~ative lenses. In the diagram of Fig. 14A, a ~chematic representation of lens surfaces similar to that representation contained in Fi~. 4B is used. ~ow-ever, arr~ws 301-304 are utilized to illustrate the deflec-tion of light at porticns of each of the optical segments of ~92~7~

each of the regularly placed lens elements. As before, the lens elements are la~eled C~, C , Al and A2.
Examining each of the elements, it can be ~een that with respect to the contiguous quadrants of each element C~, C ~ A1 and A2, all of the light impinging upon contiguous or adjoining quadrants will be directed to the s~me detector quadrant. Thus, and with respect to the lower right quadrant of element C~, the upper right quadrant of element Al, the upper left quadrant of element C and the lower left guadrant of element A2, all li~ht impinging upon these elements will be deflected to the same direction. Moreover, it will be seen that the contiguous ~uadrants together define an area the equi~alent of each of the lens elements and having its houndary described about de~lection arrow 304. This area of common deflection has been commonly shaded. All light impinging upon that shaded area will be directed to quadrant DIV of the detector.
Similarly, and with arrow 303, all liyht will be directed to quadrant DIII; and with respect to arrow 302, all light within that quadrant will be directed to quadrant DII.
Thus it can be seen ~hat from areas of the lens matrix having the same size and shape as each o~ the lens elements C+, C , Al~ and A2~ all light falling upon con-tiguous ~uadrants of the causes all light to impinge upon the same detector quadrantO
I have discovered that the detouring of light at lens elements that are of all the same power can be utilized to detect low level light image displacement. Specifically, I have found that either positive cylinder lenses7 negative cylinder lenses or astigmatic lens elements of opposite overall ~ross cyliDder alignment can be utilized to generate the optic displacement u~ilized in my invention.
~ n example of ~his utilizing a negative lens ele-ment can he illustrated with respect to Fig. 14B. Referring to Fig. 14B, a s~ries of negative lens elem~nts C- are all illustrated in ~ide by side relation. Lens elements C~ can in turn be divided into quadrants. These quadrants labeled countercloc~wi~e in accordance with the convention previously 7~L

described for detector guadrants fall into subquadrants Ql deflecting light generally to the 10:30 counterclockwise position; Q2 directing light to the 8:30 counterclockwise position; Q3 deflectin~ light ~o the 4:30 clockwise position;
and Q~ directing light to the 1:30 clockwise position.
section Ql will be directed to the detector quadrant I, all light impinging on detector segment Q2 will be directed to detector quadrant II, all light impinging upon detector segment Q3 will be directed to detector guadrant lII.
Attending to the schematic of Fig. 14B further, it can be seen that a knife edge Kl laid out on a two to one slope will haYe equal portions of the knife edge passing to all segments of the detector. For example, referring to ~ni~e edge Kl it can he ~een that egual linear portions of 1~ the k~ife edge will be deflected by each lens quadrant to a particular detector segment. For example, comparing Fig. 14B
and Fig. 15A and ex~ri n; ng the knife edge Kl from left to right, it is seen ~hat a first quarter of the knife edge will be deflected to and across detector guadrant DII. A second se~ment of knife edge Kl will be deflected to and across detector guadrant DIII; the third segment of knife edge K
will be deflected to and across detector quadrant DI,and finally th~ fourth segment of knife edge Kl to and across detector quadrant DIV. It can ~uickly be seen that egual portions of the knife edge Kl will all go to different detector guadrants.
It will be recalled from ~he foregoing discussion that two respective rules have to be followed wXen faint images are detected by the detector of my invention. The first of these rules is that when a centered image is detected, light is egually distributed ~mong all the guad-rants. The ~ecor.d rule that needs to be followed is ~hat wben displ~cement of the image occurs, ~he light impinges with a weighted impaet on the detector guadrant6. In effect an indication of the displacement of the light is given by ~he distribution of light at the particular detector guadrants.

In actual fact, this is not the case with the regular lens elements illustrated in Fig. 14B. In place and instead of ~uch a strai~ht detection of the quantity of light hitting the photodiscrete segments, I have found it necessary to differentiate between the current at certain locations as compared to the overall light signal received on all four guadrants. This aspect of the invention will be discussed more specifically hexeinafter with references to Figs.
15A-15C.
I have additionally found that by passing ~he knife ed~e over a multiplicity of elements, the criticality of the cblique alignment of ~he knife edge with respect to ~he lens matrix generated is reduced. Referring to Fig. 14C, ~uch an alignment of a knife edge is illustrated.
It will be remembered from the foregoing discussion that the knife edges when placed must follow two rules.
First, ~he edge of ~he aperture must traverse equal portions of each of the segments of the lens elements so that light from egual portions of the ~nife edge are all directed to separate detector quadrants.
Secondly, the knife edge must be disposed across the lens at a slope with respect to the boundaries of the lens elem nts and ~ot parallel to these boundaries. A par-ticularly preferred 510pe of two to one has ~een previously illustrated, khe requirement ~here ~eing present that the boundary traverse at least one set of four separate discr~te elements.
Where ~he lens elements here illustrated are laid out in a regular side-by~side pattern with rows and columns of such elements occurring, it has ~een found that placing of the knie edges in ~lignment with ~he rows and columns, or pxecisely obliquely to the xows and ~olumns re~ults in a detector configuration which will not reliably measure the displacement of ~he images.
Referring to Fig. 14C, it can be ~een that ~he knife edge can traver6e large number of discrete elements and closely approximate ~he prohibiked hori20ntal aligAment described above. Specifically, and where multitudinous '77~

elements in a side-by-side array are all created, the angle of the Xnife edge can more closely approach -the axis Qf a row ~r a column of discrete lens elements or alternately an oblique alignment of the elements without rendering the knife edge inoperative.
I have even found as illustrated Wit}l respect to Fig. 17, that the lens elements can be placed in side-by-side random alignment. With respect to such a random aligr~ent where multitudinous lens elements arP u-tilized with respect to each knife edge, I find that the distribution of light in egual proportion to each of the guadrants in accordance with the weighting of the overall image is closely approximated.
Accur~te measurement can occur with such a configuration~
Referrin~ to Fig. 15A, I illustrate a detector quadrant with knife edge illumination falling on the guadrant with respect to knife edge Kl as disposed across a lens elèment similar to that illustrated in Fig. 14B. It can be seen ~hat the respective detector quadrants are labeled counterclockwise segment DI, segment DII, segment DIII and seyment DIV. Likewise, it can be seen that the knife edge K
cuts respectively across segment DIII, DIV~ DII and DI in sequence. ~t will be noted that the detector quadrants are larger than ~he projected images from the kni~e edge. Speci-fically it is preferred if the detector area is four times the size of the image to prevent signal disparities due to image excursion beyond the photosensitive surface.
Displacement of an image in ~le X dire,ction, how ever, ~rom the con~iguration illustrated in Fig. 15A to the confi~uration illustrated in Fig. 15B produces an interesting

3~ result. Specifically, it will be immediately observed that with displacemen* merely in the X axis direction, ~he amount ~f knife edge in detector ~egments DI plus DII or DIII plus DIV rF~m~; n~ unchanged . ~owever, this is not the case with respect to detector segment5 DI plus DIV or DI I plus DI I I -35 For ex~nple, the length of knife edge Kl in detector segmentDI 1~ is reduced . This kni~e edge segment appears instead at ~egment DIV.

~;27~

Displacement of the image in the Y direction from the configuration illustrated in Fig. 15A to the configura-tion illustrated in Fi~. 15C likewl6e produces an interesting result. Specifically, it will be observed that with dis-placement merely in the Y axis direction, the amount of knifeedge in dete~tor segments DII plus DIlI or DI plus DIV
remains unchanged. ~owever, this is not the case with respect to detector segments DI plus DII or DIII P IV
at the amount of light in each guadrant during the motion from the configuration in Fig. 15A to the position of Fig.
15C does produce some non-linearity. First, and during the first part of the motion, it will be seen that the amount of knife edge in guadrant DII reduces until all of the knife edge Kl passes out ôf quadrant DII. When this motio~ has occurred, the knife edge will ~hen pass out of the detector ~uadrant DI. There will be at detector quadrant DII no further light reduction. In short, there is a non-linearity resulting from the displacement in the Y direction for each quadrant seen separately, but the sums of DI plus DII or DIII
plus DIV behave in a linear fashion with translational motion in Y.
I have found ~hat by differentiating the sums of total light received with respect to the light received at cextain guadrants, a signal proportional to the displacement in the X and Y directions can be generated. For example, where displacement occurs in ~he X direction, I find ~hat by the following formula a signal with respect to displacement in the X direction can be generated:
3~
D - LI-LII-L~ LIV
~ LI~LII+L~ LIV

Similarly, because of the non linearity appearing in displacement ~long the Y a~is as illustrated in Fig. 15C, I again have found that by differentiating certain of the ~egments with respect to ~he other detector segments in comparison to the total light received, a signal with xespect ~Z7~

to the Y axis displacement can be generated. Such a dis-placement can be obtained by the formula:

D - LI~LIl-LlII LIV
y ~ LII~LIII~LIV

where:
~X is the displacement in the X direction;
Dy is ~he displacement in the Y direction;
LI is the light impinging upon quadrant I;
LII is ~he light impinging upon guadrant II;
L~II is ~he light impinging upon ~ladrant III; and, LIV is ~he light impinging upon detector guadrant XV.
I~ the use of most objective xefractors, ~here is a problem of positioning which is commonly encountered~ Speci-fically, the eye must be acguired. Acquisition includes placing the eye in the proper alignment to the optical axis of the instrument or in what may be ~escribed as a "XY"
positioning. Moreover, once the eye has been acguixed along ~he optical axis, the towards and awa~ position of the eye is import~nt. ~or this aspect o~ the invention, a specialized aperture has been developed.
Referring to Fig. 16A, a detector I had utilized with this invention is illustrated. Specifically, four prisms 401, 402, 403, 404 are placed in a sguare array. The prisms placed in their ~quare array define a central sguare aperture 410 and ~our peripheral ~guare apertures 411, 412, 413 and 414. Each prism has an opaque face and three beveled edges from which light is emitted. In the case of prism 401, there is an opaque face 400 and three light emitting edges 415, 416 and ~17.
Each ~f the respectiYe edges has a light emitting diode focused through a lens. The light emitting diode is focused through a lens and thence through the prism so ~hat a yreatly enlarged image of the light emitting diode is focused at the eye to be examined. In the case of prism 401, light emitting diode 405 is focused through lens 409 and has two refractions and ~ne reflection from and within prism 401.
These light deflec~ions cause the light to be emitted from 5 prism edge 415. Typically, the beveled edge of prism 415 is aligned so that the focused light emitting diode is directed to and upon the eye. Preferably, a "pebble plate" surface is added to the prism optics, preferably at the surfac~ of first incidence of light into the prism.
Similarly, light emitting ~iode 406 focuses through edge 416, and a light emitting diode 407 focuses through edge 4170 It will be undeLstood that each of the respective prisms 402, 403, and 404 h~ve a light emitting edge similar to those of prism 401.
All ~nife ed~es are preferably masked so that light inGident immediately over them are passed to the detector and the remainder of the light is rejected. This masking is illustrated in the view of Fig. 16A.
It will be noted that the corners of the light emitting edges are masked. For instance in the case of prisms 401 and 402, it will be seen that the corners 42~ are covered.
From the respective prisms, light is emitted to the eye to be examined, and returns from ~he eye being ~xamined by way of projection optics which have been previously illus-trated and are not sho~n here. The received light passes over the knife edge defined by ~he junction of the prisms and the apertures. The light then passes interi.orly of a detec-~or having the square aperture array previously illustrated.
~hen passing interior of the projector, the liyht passes ~hrough the specialized lens element V ~preferably khe pebble plate illustrated hereafter in ~ig. 17) and thence through focusing lens L to ~he detector D where an image ~" is formed. Analysis of a knife edge image occurs.
Referring to Fig. 16B, a view of the imagin~ appa-ratus along line 16b of Fig. 16A i~ illustrated. Specifi~
cally, ~he detector is ~hown 50 that the light emitting edges may be viewed as they are ~een from the eye of the patient being examined.

~;Z 7~4 It will be noted that the lighL emitting edges 416 on one hand and 418 and 419 on the other hand are disposed along a top colinear horizontal edge of the dete~to.r. Edge 416 is equal to the lengths of edges 418 and 419 added to-5 gether. Thus it may be fairly said that the two outsideedges when added together have the same length as the inside edge 416.
It will be al50 noted that edge 416 poin~s in opposite direction from edges 418 and 419. Thus, assuming that the edge comprising edge 416 facing in one direction and edges 418 and 419 facing in the opposite direction are illu-minated, an eye will have equal and opposite refractive effects produced ~herein by the various edges. This is another way of saying that the edge effects will not comprise a weighted image giving a tell-tale indication of either spherical or cylindrical correction being required. In other words, illumination along a single edge with equal lengths in opposite direction will produce no detectable prescriptive correction.
Referring to the linea. edge comprising the illu-minated edges 426, 428 and 429, the same statement can be made. Since equal lengths of edge are illuminated in oppo-site directions, weightin~ of the images in the eye will not be detected. It can be shown, however, with respect to Fig~
~5 16B that the ~equential illumination of these re~pective images can serve to assist to position an eye.
Referring to Fig. 16~, a schematic di~gram is ~herein shown. The schematic assumes that ~he eye is illus-trated properly centered in the X and Y plane. Naturally, by measuring the imag~ impin~ements on the quadrants of a detec-tor DI, DIX/ DII~, DI~, centering of the eye with respect to an optic axis ca~ ~ccur.
The question then becomes what is the proper posi-tloning of ~he eye in the Z axis direction.
In ~he schematic of Fig. 16C, ~he respective light emitting edges are ~chematically sho~n. 5pecifically, edges 416, 41B, and 419 are all illustrated. Similarlyt lower edges 426, 428 and 429 are all illustrated.

It should be realized that Fig. 16C is a schematic.
Focusing optics P schematically illustrate the convergence of ~he irnage from the edges to ar active detector. The ~pc~cial-ized optics V AS well as the eye of the patient are all omitted.
In Fi~. 16C, the images for each of the knife edges at differing distances are illustrated. Referring to 'che si~
detector images shown, the upper two images are for when the eye is at the proper distance from the detector. The middle image is an illustration of the detector when the eye is too close. The lower two pair of detector images are illustra-tions where the eye is too far away.
It will be under~tood that the right-hand yroup of images are the image ~hat would be cast where knife ed~es 418, 416 and 41~ are illuminated. The left-hand group of images are where edges 428, 426 and 429 are illuminated.
Typically, these images would be produced with flrst one linear set of knife edges being illuminated and thereafter a second linear set oX knife edges being illuminated.
Referring to the upper images where the eye is positi~ned the proper distance from the detector, it can be seen that the image formed by knife edges 418, 416 and 419 are the same as the images being formed by knife edges 428, 426 and 429.
2~ Where eye is too close, the images formed by knife edges 418, 416 and 419 raise up ~n the surface of the detec-tor. Great concentrations of resultant images appear at upper guadrants DI and DII. The effect on the image of knife edges 42~, 426 and 429 is ~he opposite. Specifically, ~he respectiYe images of the knife edges fall in greater measure 0~ guadrantS ~III a~d ~IV
Typically, ~he knife edges of the detectors are either modulated with their own discrete ~i~nal ~o that the images can be separated one from another, or are alternately illuminated. In either ca~e, the resultant weighting of the detector signal at ~he guadrants ~ he detector gives an indicatlon of the toward~ and away position of the eye (not shown).

7'7~

As can be seen in the lower illustration, where the eye is too far away, the effects are reversed. Specifically, for ~nife edges 418, 416 and 419 the lm~ge shifts downwardly.
Specifically, the image shifts to detector ~uadrants DIII and DIV-For the knife edge image of knife edges 42~, 426and 429, the effect is reversed. The knife edge shifts upwardly to detector quadrants DI and DI~.
It will be observed that ~he particular knife edge images cast are symmetricalO That is to say, they are egual-ly weighted about a center line. This is because the knife edge images oppose one ano~her for equal lengths. Conse~
quently, it will be appreciat~d that ~e particular knife edgé images cast are insensitive to the particular optical prescription ~hat may be encountered in the eye.
Thus, it can be seen that the image produced is insensitive to the prescriptive effects the eye might have but is sensitive to the positional effects that the eye imparts in being acquired by the instrument.
Assuming that the eye is properly acguired, the measurement of the eye then occurs by illuminating light knife edges dispcsed along the same direction but at varying positions. A knife ~dge examination utilizing only one such group of knife ed~es will be illustrated, the knife edge examination of other edges being analogous and easily understood.
Referring to ~he schematic of Fig. lSD, a typical knife edge test is illustrated. Specifically, ~nife edges 416, 4~8 and 429 are all illustrated. The knife edges are 30 illustrated passing throuyh projection optics P to a detector consisting of detector guadrants DI ~ DI I ~ DI 1 1[ ~ and DIV .
First, it will be noted that all of the knife edges 416, 428 and 429 are addressed in the same direction. As they are addressed in khe same direction, the xesultant image produ~ed by an eye will be ~nife edge sensitlve as to ~he prescriptive coxrection reguired. This being ~he case, and assuming ~hat we have an emmetrope, the detector seymentS
illustrated will be a ~i ni mal imaqe~ As ~le respecti~e knife ~277~L

edges are spaced evenly ~bout the central axis of the optic instrument so as to produce a centroid of ill~nination evenly about the Optlc axis of the instrument, the measurement system will have its position sensitivity minimized. That is S to say, its position sensitivity to the positioning of the eye wi~hin the instrument would be mlnimized.
In accordance with the previous illustrations rendered, the hype~metrope ~ill produce an image on one side of the detector, ~y detector quadrants DI, DII. Similarly, the myope will produce an image on the opposing ~uadrants DIII, DIV. Finally, an astigmat will have an image on the quadrants on one side or the other side, the image here being ~hown on guadrants DII, D~
As will be reali~ed by those having skill in the art, the edges of the detector can be ~witched. They can be switched so ~hat images opposed to those illustxated can next be taken. This gives the instrument the desired push~pull effect. Moreover, it can also be realized that the imaging can be accomplished left and right. That is to say, a mea-surement can be taken using a group of edges on the left andthen an opposing group of images on the right.
It will be realized at this point that the light emitting diodes can be modulated as can the detectors util-ized with them. Specifically, the measurements can all be taken simultaneously with the modulated signals reoeived back from the eye segregated. Moreover, by using a central and visible target for fixation, foGusing o~ ~he eye to a visual target may result. This focusing of ~he eye cah ~hen have the disclosed objective refractlon superimpQsed thereon.
As to the particular imaging scheme chosen, it ~hould be understood that the edges are all active and given a common centroid. Thus when they fall upon ~he detector D, they fall upon each of ~he ~uadrants with equal intensity.
Referring to the view of the optical train shown in Fig. 6E
and the corresponding image of the detector ~hown in Fig. 6F, ~he balancing of ~he ~pecular re1ection image with respect to the alignment of detector~ utilized to measure the pre-~criptive effects of the light is illustrated.
!

Referring to Fig. 16E, an eye E has three sources A, B, C imaged thereon. Images of these sources are relayed by optics (no~ shown) to three real image locations. These image locations are KA, ~, Cc.
Image KA is above the optical axis and twice as long as respective images ~ and Kc. An image of these respective optics is relayed through the specialized optics V
to the detector D. Specialized optics V has bee~ previously described.
Referring to Fig~ 16F, the centroid of light on the detector D is illustrated. This centroid is for specularly reflected light and does not incorporate any prescriptive corrections.
It can be seen that each image is off-set from the optical axis. Specifically, it is off-set by a given amount.
Thus, lf the detector D is either too close or too far away, the respective movements of the image ~rom each of the light sources will remain ~he same.
Referring to Figs. 16G and 16H, it can be seen that this is not ~he case where a single knife edge is utilized.
In Fig. 16G, a pupil with a single light source A has the image ~lereof broadcast onto a specialized optical plate B at the illustration knife edge KA. The knife edge KA is there-fore relayed by ~ptics not ~hown to the detector plane.
Assuming that the detector pl~ne is at khe right distance from the eye, the image will impinge upon ~he cen-ter. However, if the eye is Pither too far away. or too close, the image will move. Specifically, it will move off center. In Fiq. 16G, the image of a pupil mGved away from the center of ~he eye is shown.
Referring to ~ig. 16H, an on~enter image is illus-trated. It can be seen ~hat the light centroid is off-center with respect to ~he detectox quadrants DI, DII, DIII, and D~v~ In ackual fact, the migxation of the image has occurred from ~he two upper guadr~nts DI~ DII to and towards low2r quadrants DIII~ DIV-Returning ko ~he three dekector array shown in Fig.3 and taking the case of ~he non-specularly reflected light, the action of the towards and away positioning of the optics here illustrated can be illustxated.
Specifically, and if detector D is at the position Dl with respect to specialized optics V and the images KA, and Kc, it will be seen that all images will be broadcast into suhstantially coincidence. That is to say, ~hey will be imaged upon a central point of the detector ~
If, however, the detectox is too far away such as at position D~, three such images will result. ~hese three such images are illustrated in Yig. 16L.
Referring to Fig. 16L, and taking the case of myope, it can be seen that the three images are produced.
The lower image IA will be twice as intense as the two upper images IB and Ic. These ima~es IB and IC will all be dis-~5 placed in accorda~~e with the particular prescriptive cor-rection of the eye being required. This being the case, and reviewing thP images heretofore discussed, it will be seen that the displacements will add in all detector guadrants DI ~ DIV to give ~he same result as the single image shown in Fig. 16K. Conseguently, it will be realized that the detec-tor scheme herein illustrated is insensitive to towards and away positioniny of the eye with respect to the apparatus.
It will be understood that with this explanation an immediate process can be added. First, axial towards and away alignment such as ~hat illustrated wi~h Fig. 16C will be undertaken. Thereafter, and once the eye is grossly in place, prescriptive measurements will be made. These mea surements will be made by apparatus illustrated in accordance with Figs. 16J, 16K and 16L. Thus, even though once the eye is properly positioned and the eye ~anders somewhat rom its original positionin~, the disclosed optics will be relatively in~ensitive to such movement. Correct objective refraction will result.
Regarding specular reflection, and referrin~ to the view of Fig. 16F, it can be ~een ~hat the areas o~ thP light ~ources are important. Specifically, by having the moment of optical areas ~he same above and below the horizontal axes as well as the moment left and right of the vertical axes being ~ 2~

the same, specular reflection from the eye will cancel itself among the various detector segments. Consequently and with the edge arrangement sho~n, perturbation of the refr~ctive findings by return specular reflection car~ot occur.
Referring to Fig. 16J,'an alternate dimension of ~he knife edge configuratiorl is illustrated. Specifically, each of the knife edges Ka, Kb, Kc are of the same length and area. These respective knife edges are separated from a horizontal axes by two units of distance in the case of the knife e~ge Ka and one unit of distance in the case of ~he knife edges Kb~ K~o The unit of distance are all labeled with 2a for knife edge Ka and la for knife edges K~, K~. The knife edges are all of the same length~ Specifically, the knife edges are labeled with the width dimension D3.
Referring to Fig. 16L, the unfocused centroids of the image are there shown. Specifically, it can be seen that the lower ima~e Ia is displaced from the horizontal axes by an amount appxoximately twice the centroid of the two upper knife edge ima~es Ib~ Ic. Perturbation of the refractive signal due to axial or towards and away displacement will not occur. It should be pointed out that for best performance, the light receiving or viewing apertures adjacent to knife edges should also have substantially egual moments above and below the horizontal axis as well as left and right of the vertical ~xis.
Turning attention to Figs. 18A~D, these figures illustrate ~he patterns which form on the detector due to a decentered pupil with an arbitrary refractive error (sphere plus cylinder at a tilted axis to the knife edge).
Fi~s. 18A and 18B illustrate horizontal knife ~dge interrogation. The ~nife edge K in Fig. 18A is disposed so that light passes to the receiving area 40D below the ~nife edge K and over the linear boundary 415. Likewise, in Fig.
18B, an area 402 receives liyht immediately ~bove ~he knife 3~ edge 415. With respect to ~ig. 18C and 18D, the knife edges are vertically disposed. I~he ~dges there respec~ively are to ~he left of and to ~he right of the detector surfaces. Areas 404 and 406 receive ligh~ in Figs. 18C and 18D respectively.
I

~92, ~L

Each of the Figures 18A-18D has schematically illustrated next to the respective knife edges the detector ~ur~ace. The detector is that detector illu~trated previously.
In the case of the imaye illustrations herein given, it will be understood that the light i5 distxibuted to the detector plane by the preferred optics shown herein.
Thus, the light recei~ed at ~he detector plane will not have the appearance schematically illustrated on the detector surfaces of Fig. 18A~]~D. Instead, the light will be ev~nly distributed among the detector ~uadrants as previously set forth.
In each case of Figs. 18A D, the detector measures two values which are proportional to the X centroid position times the total received light flux and the Y centroid posi-tion times the total received light flux. Since the total fIux is the same for both values, the valu~s are in fact proportional to the X and Y centroid positions.
In addition, it will be appreciated the source and detector array are designed 80 that each knife edge has equal values for total light and in fact is symmetrical in all respects about the pupil image center on the detector. Thus, the measured values can be added and subtracted in a method which will now be given so that both refractive information and pupil decentration information can be e~tracted.
Note in Fiy- 18A~ ~CA ~ ~ ~ ~
YCA ~ A ~ YP
where XcA - X centroid position YCA = Y centroid position - X displacement of centr~id from pupil center A = Y displacement of centroid from pupil center ~ - X position of pupil center Yp ~ Y position of pupil center Similarly and in Figure 18B, XcB ~
YCB ~ ~ YP

Due to the pattern symmetry set forth above, Rx~3 RXA
A
50; X
YC~ yl This means then;
~ l XC~ - Xp t R~ t Xp -RX~ P
measured values YCA ~CB YP ~ A ~ A ~P ~A ZYP
This 5hows that the measured values can be added, X to X, Y
to Y, to yield values which are directly proportional ko pupil decentration. Note th~t prescriptive information is not included.
Likewi se:
~CA X~B = ~ ~ ~ -(~ - ~ )=2 ~
YCA YCB YP ~ A (YP ~ A~ 2 ~A
which shows that a correct subtraction of measured values yields values which are directly proportional to the dis-placement of the centroid of the received pupi]. pattern from 2U the pupil center. In addition, because these values are X
and Y displacements of the centroid, ~hey yield both magni-tude and direction of ~his displacement which in turn are directly r~lated to refractive error as previously set forth at length in ~his application.
It has heretofore been mentioned ~hat, in this application, one parallel set of knife edges c2nnot provide - complete refractive information (although it does give decentration of the pupil). ~owever khe remaining informa-tion is collected via the second parallel set of knife edges as ~hown in Figures 18C and D. Note that in all figures the relative position of the pupil center to detector center is ~he ~ame.
In ~ ~ry, by adding all X centroid values a value proportional to X pupil decentration is obtained. By adding all Y centroid values, a value propGrtional to Y pupil de-centration is obt~ined. By correctly subtracting values of parallel knife edge pairs, four refractive proportional values arise, namely;

, 7~
XCA - XCB + 2 RXA

CA YCB YA
X~c - XCD 2Rxc CC CD 2Ryc Then it is found tha-t values proportional to sphere equivalent tSeg), cross-cylinder axis 90/180 (C~) and cross~
cyllnder axis 45~/135 (Cx) can be obtained by combining the refractive proportional values in the following manner:
eg XC ~A
C+ ~R~ A

X XA RYC
where C~ is 0-90~ cylinder, and CX is 45-135 cylinder~
It will be apprec.iated that the detector disclosed herein can be utilized to have refracting optics driven so as to null the received signals at the detector surface~ I have demonstrated such circuitry before in my prior U.S. Patent 4pO70,115 issued Januaxy 24, 1978. Specifically, that paten~
disclosed an invention which may be abstracted and summarize~
as follows:
A lens meter is disclosed in which continuously variable spherical and astigmatic corrective optics are manipulated to measure the prescription o~ a suspect optical system. A target including a straight line is focused for maximum clarity, the target being arbitrarily aligned without respect to the a~is of the suspect optical sys-tem. Continuously variabl.e sphexical ~2 o 7~
and first astigmatic optics are juxtaposed to the suspect optics and the image o~ the -target projected through both the suspect optics and the conti.nuou~l~
variable optics. Spherical and first astigmatic corrections alony at leask one axis diagonal to the line target is made until maximum sharpness of a projecteA image oE ~.he line results. A first component of astigmatic correction results. A second target, again consisting o~ a straight line, is introduced~
this target is angularly ~8 inclined with respect to the first target preferably a~
45O Spherical adjustme~t is made together with a diagonally aligned secon~ astigmatic correction along at least one axis diagonal to the second line taryet until maximum sharpness of the projected image of the line results. A second component of asti~m~tic correction and final spher:ical correction results. Provision is made for remote mani.pulation of the continuously vari-able opkics to determine prescription automatically.
A ~epresentative clalm of that patent is included as follows:
1. A process for measuring power of a suspPct optical ~ystem in at least one component of cylinder including ~he steps of: mounting said suspect op'cical system in 3 light path; projecting light including an image of at least one first straight line target of first arbitrary preselected an~ular alignment without regard to any suspected principal axis of the suspect optical system along said light path; providing in said ~0 light path variable optics for movement to a power of sphere and cyli~der substantially equal and opposite to compo~ents of sphere and cyliIIder in said suspect op tics, said variable optics including variable spherical optics to vary the spherical component of light pro-jected there through and variable cylinder optics for varying khe astigmatic lens power along first intersect-ing diagonals at substantially equal and opposite angles from ~he preselected angular alignment of said first ~traight line target; projecting an image u~ said ~traight line target from said light passing through ~aid variable optics and said suspect optics; and, varying said spherical optics and said first astigmatic optics to optimize the image of said projected straight line target.
Referring to ~hat ~aten-t, at Fig. 5, sufficien1 ~chematic circuitry is given from a detector having four distinct quadrants to drive optics to achieve a null image.
While adaptations must of necessity be made to produce the 7~

detector configurations herein set forth, it is believed that ~uch chanyes may be easily be made by those having ordinary ~kill in the art. L,enses ~chematically achleving such a null image are shown in Fig. 16G as variable spherical lens 516, 0~ 90 cylinderical lenses 578 and 45-135~ cylindrical lenses 520. These lenses are taken directly from Fig. S of the referred to by reference patent.
It i~ a particular advantage of my invention that xefractive information returneA rom the eye is not dependent upon the ability of khe eye to return light to the detector.
Take the case wherein a retina, through disease, has enlar~ed blood vessels, and/or other configuration. Consequently the xetina i~ not capable of uniforml~ returniny light to the detector over its surface. III such cases, the light received back by one of the knife edges in Figs. 18A-18D will substan-tially differ from ~he light received by other knife edges.
By the expedient of mathematically equating all of the returned light - giving the guantity of returned light in each knife edye alignment of Figs. 18A-18D the same value and ~hereafter processing the values, the effects of irregulari-ties in the retina may be ignored.
It will be noted that in the previous description and eguations relative to Figs. 18A-18D, I haYe effectively illustrated ~'moments" of the light flux with respect to the particular detectox guadrants utilized. Thus, when ~he term "moments" is used heretofore or hereafter in this applcation, it should be so understood.
It will be understood further that for the bes-t performance, the apertures her~in utilized should be symmetri 30 cal. Moreover, the areas of the apertures and the receiving areas should all have equal moments.
Although the poink has heretofore been made, it should be emphasized that in the case of the knif~ edges, disposition at right angles is not required. For example, the knife edges could be disposed at 45 anqles. M~reovex, and with variations to the mathematics herein disclosed, a~d/or optics detector ~urfaces or bo~h, varying angles could be used between the interrogating knife edges. I have merely illustrated the preferred parallel and opposed knife edges in symmetrical alignment to set forth the preferred er~odiment of my invention as known to me as this moment.

It will be understood that the disclosed invention 5 will admit o a number of er~odirnents. For example, any projection system between the disclosed wobble plate and eye may be utiliæed.

:

Claims (4)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. An apparatus for detecting low level patterns returning from the eye for testing the eye comprising in combination: a detector, said detector including a plurality of apertures; knife edges aligned to at least some of the boundaries of said apertures and terminating inwardly along straight lines defining a view path to and towards said aperture over said knife edges; at least two of said knife edges facing in opposite directions across said central aperture.

2. The apparatus of claim 1 and wherein said knife edge includes a central aperture and four peripheral apertures.

3. The apparatus of claim 1 and wherein said aperture transmits light only immediately over said knife edges.

4. The apparatus of claim 2 and wherein the dimensions of said central aperture are greater than the dimensions of each of said peripheral apertures by a factor of 2.