GB2161600A - An optical height measuring system for operation in a noisy environment - Google Patents

Abstract

An optical height measuring device 10 includes an array 25 of photosensitive cells to which light reflected from an object 15 is directed. The size of the spot of light is controlled by optical system 26 to be greater than one cell size d and less than 2d, whereby the light impinges on at least two adjacent cells. Object height is determined based on the outputs of a pair of two adjacent cells selected from the array. Signal processing means includes means for generating for different pairs of adjacent cells i + j, corresponding signals <IMAGE> where Ci and Cj are outputs of adjacent cells, and means for selecting from these signals any signal whose value lies between certain limits and using their value(s) to determine the object height. <IMAGE>

Description

SPECIFICATION An optical height measuring system for operation in a noisy environment The present invention is generally related to an optical detector and, more particularly, to a high resolution optical detector with high noise immunity.

In the art of automatic welding by means of a robot, the heights of incremental areas or pixels along the weld seam have to be determined in order to produce a proper weld. The height of each pixel is typically determined by an optical assembly which includes a source of light, e.g. a laser, whose beam is directed to the pixel. The height is determined based on the location of an image of the light, reflected by the pixel, on the surface of an array of light sensitive cells. The array is typically referred to as the optical detector or simply the detector.

It is apparent that for proper operation high height resolution is required. Also the height indicating signal should be quite high. With presently used detectors neither or both of these requirements are not met. Cells on the order of 1 OOzm X 200 ym provide high signals but low resolution, while CCD's provide high resolution but low signals.

Also, in welding operation considerable noise is present. It includes frequencies within the laser band. Thus in spite of filtering, noise greatly affects the detector's output. That is, the indicated height is subject to noise. Furthermore, when welding, substantial smoke is produced.

It frequently affects the size of the image projected on the detector and thus its output, which is most undesirable. It is to solve these problems that the present invention is directed.

The present invention may be defined as an arrangement comprising: means for directing a beam of light to an element; and optical detector means for receiving light reflected thereto by said element to determine height element as a function of a preselected functional relationship of the outputs of cells of an array of cells forming part of said optical detector means.

Briefly, in accordance with the invention an optical detector, consisting of a linear array of a given number of cells, e.g. 16 is used. Optics are included so that the size of the image which is reflected onto the cells is greater than one cell dimension and less than two cells. The outputs of adjacent cells are processed to generate certain functions, and the height of the element is determined based on one of these functions, when it is between certain function limits.

Let it be assumed that the image size to be w and that of each cell to be d, where d < w < 2d.

Designating the outputs of adjacent cells as Cj and C the function which is generated for each pair of cells is

It is one of the values of these functions which is used to indicate height.

In accordance with another aspect of the invention optics is provided to insure that the size of the image, directed to the optical detector remains on the order of w even though the light may pass through a window, made partially opaque by smoke, which is created as a result of the welding operation.

The novel features of the invention are set forth with particularity in the appended claims. The invention will best be understood from the following description when read in conjunction with the accompanying drawings: Figure 1 is a simple isometric diagram useful in explaining one application of the present invention; Figure 2 is a simple diagram for explaining the use of an optical detector for height measurements; Figures 3 and 4 are diagrams useful in explaining the basic aspects of the invention; and Figures 5a and 5b are optical diagrams useful in explaining a novel optical arrangement of the present invention.

Attention is directed to Fig. 1 wherein numeral 10 designates an optical system, located ahead of a welding robot 12, which produces a welding arc 14. The function of system 10 is to locate the seam 15 between pieces 16 and 17 which are to be welded together, as shown at 18. The location of the seam 15 is provided by determining the height of incremental areas of pixels thereof as the system is advanced along the seam and scans across it, to determine the lowest point of each pixel.

The optical system 10 is shown in Fig. 2, wherein a light source, e.g. a laser 22 directs light to the seam from which light is reflected to an optical multicell detector 25, through an optical system, represented simply by lens 26. As is appreciated by those familiar with the art the position on the detector surface whereat light impinges is related to the height of the surface from which light is reflected to the detector.

In accordance with a preferred embodiment of the present invention the detector 25 consists of a linear array of a plurality of photocells. For example, it consists of sixteen photocells, designated in Fig. 3 by C1-C16 from left to right. For explanatory purposes let it be assumed that light directed to cell C1 represents minimal height while light directed to cell C16 is indicative of maximum height. Let it further be assumed that the size (width) of each cell is equal to and is designated by d.

In accordance with the present invention the image of the light which is directed to the detector 25 is shaped by optics 26 so that its size, designated w is always greater than 1 d and less than 2d, i.e. d < w < 2d. Consequently, at any time the light image, designated by 30, and also referred to hereinafter as the light spot or spot, always illuminates more than one cell. As shown, the output of each cell is supplied to a signal processor 35. Clearly the magnitude of the output of any cell depends on how much of its surface is illuminated by spot 30. For explanatory purposes total cell illumination is assumed to be equal to unity or one (1) while zero illumination results in zero (0) output, ignoring noise.

Unlike the prior art herein, at least parts of two cells are illuminated by spot 30.

In accordance with the present invention processor 35 processes the outputs of the cells to provide a height indication output, designated by 40 which is a function of the outputs of two adjacent cells which are illuminated simultaneously. Briefly, processor 35 generates a function f for each pair of adjacent cells, e.g. C1 and C2, C2 and C3, etc., where

and wherein Ci and C represent the outputs of cells C1 and Cj.

Assuming w to be equal to 3/2d and that only half of C2 is illuminated while all of C3 is illuminated 0.5-1 0.5 f2 = ~~~~~~ = - - = -0.33, 0.5 + 1 1.5 representing a specific height. If the height were to increase slightly (spot 30 moving to the right) less of C2 would be illuminated and part of C4 will become illuminated. For example, if 4/10 of C2, all of C3 and 1/10 of C4 are illuminated processor generates for C2 and C3 and for C3 and C4 the following functions.

0.4-1 0.6 f23 = ~~~~~~~ = - - - - .428 and 0.4+1 1.4 10.1 0.9 f3,4 = ~~~~~~- - =.818.

1 +0.1 1.1 Similarly when the output of each of cels C2 and C4 is 0.25 and that of C3 is unity 0.25 - 1 0.75 f2,3= = = - ~~~~ = -0.6 0.25 + 1 1,25 while 1 -0.25 0.75 f3.4= = = ~~~~ = + +0.6.

1+1.25 1.25 From the foregoing it should thus be apparent that a slight change in height, resulting in a small movement of spot 30, results in a significant change in the function value. In some cases for a given spot position, representing a specific height, two adjacent functions have different values, e.g. f23 = - 0.6 and f34 = + 0.6.

It should be appreciated that each function is essentially S shaped with a central substantially linear portion in which the change in function value is more proportional to change in height.

Also, as apparent from above, when the spot is positioned so that it fully illuminates one cell and equally illuminates adjacent cells on opposite sides the values of the two adjacent functions, representing the same height, are equal but of opposite sign, e.g. f2 = - 0.6 and F34 = + 0.6.

These properties are used by processor 35 to choose the function whose value is then used to provide the height indicating signal.

Briefly, the detector is first calibrated by moving the spot 30 from one end of the detector array to the other, as if the spot is reflected from different heights. For each spot position the functions are generated and the value of the function in its linear portion is recorded. At the end of the linear range of one function the corresponding value of the adjacent function for the same spot position is used. For example, when f2 = - 0.6 representing a given height, a switch is made to function f34 = + 0.6 which represents the same height. This process continues until for each height a specific value of a function whtin its linear range has been determined.

This operation may be described in connection with Fig. 4 wherein only three functions f2,3, f3 4 and f4 5 are diagrammed with the crossovers being when adjacent functions have values of - 0.6 and + 0.6. The vertical direction of each function represents its values while corresponding heights are along the horizontal direction.

In operation, to determine the unknown height of a pixel along a seam or the like, the output of the cells Cl-Cl 6 are first processed to generate the functions f12, f22, etc. Thereafter a determination is made which function produces an output which lies within the function's linear range, e.g. + 0.6 to - 0.6. Then the output is used to locate a height value derived during initial calibration for that value of that function. Successive outputs of the same function are used to derive height values based on the generated values of the function. Then when the output of the function reaches one of its linear limits, e.g. - 0.6, values of the next function are used to determine successive height values.

It should be appreciated that since height values are determined not as a function of light on one cell but rather as a function of light on at least two cells, the detector of the present invention is highly immune to noise. This is due to the fact that noise affecting ond one cell equally affects an adjacent cell and the difference gets cancelled out. Also, since herein the output is only a function of two cells out of many, e.g. 16, the noise affecting the outputs of the other cells does not affect the output of the novel detector of this invention. The optical detector of the present invention can thus operate successfully and accurately in a noisy environment.

In the foregoing the functions to be performed by signal processor 35 have been clearly described. It should be apparent that it can be implemented with discrete circuits to perform the following: 1. generate the functions fj.j e.g. f1,2, f23, etc.

2. determine the function whose value is between predefined upper and lower limits, e.g.

+ 0.6 and - 0.6; 3. use the value of the selected function to determine the height value, which such value represents based on the precalibration; 4. Switch to use values of an adjacent function when the value of a function, having been previously used, reaches its linear limit, e.g. the value is equal or greater than + 0.6 or is equal or smaller than - 0.6.

Since circuits capable of performing any and all of these operations are clearly within the knowledge of those familiar with the art, they need not be described in any detail. It should also be clear that all of these operations may be performed by a computer, programmed based on the functions which are to be performed and which have been described. Again, based on the description, different programs can be written based on the computer to be used. The foregoing description is clearly sufficient to enable any competent programmer to write or compose the program without having to resort to any inventive steps.

It should be apparent that the novel optical detector is based on the fact that the size of the light image or spot is w, which is greater than one cell size d and less than 2d. Typically, the light which is reflected to the optical system 26 (see Fig. 2) from the workpiece passes through a transparent window designed in Fig. 2 by line 50. The function of window 50 is to pass the light only, yet block off any polluted air, e.g. smoke, created by the welding operation. Due to such smoke or the like a danger exists that the window may become partially opaque and thus reduce the size of the light passing through it. This in turn would affect the size of spot 30, and may reduce it to be less than desired. Alternatively stated the size of the spot 30 may decrease to equal d and even less, which would affect the entire operation of the detector.Furthermore, a shift in the spot position may take place which would result in an erroneous height indication.

This problem has been solved by a novel optical arrangement 26 which in Fig. 2 is represented for simplicity purposes only by lens 26. The novel optical arrangement 26 may best be described in connection with Figs. 5a and 5b to which reference is now made.

Fig. 5a represents a conventional optical arrangement in which a light source 60 of a given size passes through window 50 to a lens 62. By proper choice of the lens and the distances an image 65 of a predeterminable size is formed of source 60. The intensity of the image is typically gaussian or bell shaped as shown in Fig. 5a and designated by 66. The intensity is symmetrical about the axis 70 of the optical arrangement.

If however window 50 were partly opaque, as represented in Fig. 5a by 75 so as to block off part, e.g. the lower half of source 60, the image size would not be the same as herebefore described. Rather, the size of image 65 would be one half with an intensity distribution as that designated by 66a. Although the peaks of the intensities are about equal, that of 66a is offset by Ah from axis 70. Such offset represents error in height read-out due to the effect of smoke or the like on window 50. It is thus clear that conventional straight-forward optical arrangements cannot be used, where light has to pass a window capable of being fogged up or made opaque over a portion thereof.

Attention is now directed to Fig. 5b in which the solution provided by the present invention is diagrammed. In accordance with the invention the lens arrangement 26 includes a unique optical element 80, and a plano-concave lens 62x, located fore of element 80.

The flat side 82 of element 80 is juxtaposed the plane side of lens 62x, while the oppsite side of element 80 defines a plurality of small cylindrical plano-convex lenses 85. The light passing through these small lenses are made to converge to form the image 65. The contribution of these lenses to the image 65 is such that even though part of the window is blocked off, such as by smoke 75, the size of the image 65 remains substantially constant. The image intensity is that of a mesa with a plateau of uniform height with two steep narrow sides. Unlike the arrangement shown in Fig. 5a, in the novel arrangement of Fig. 5b the fogging of window 50 i.e. the reduction in the size of the source 60 is only a reduction in height of the intensity of image 65 without affecting its size.In Fig. 5b the intensities of the image without a fogged up window and when partially fogged up, are designated by 90 and 91 respectively.

In one embodiment, actually reduced to practice, a lens 62x with f = 40 and with an element 80 with cylindrical lenses of a width d = 0.55 mm and a radius R = 50 mm or f = - 100, a mesa shaped image go of a plateau width of 200 m with sloping sides of 25 itm each, was produced at a distance of 45 mm from the plane side of lens 62x, spaced 9 mm from the flat side 82 of element 80. Any partial obstruction of window 50 only affected the height of the image but not its width.

Although particular embodiments of the invention have been described and illustrated herein, it is recognized that modifications and variations may readily occur to those skilled in the art and consequently, it is intended that the claims be interpreted to cover such modifications and equivalents.

Claims (11)

1. An optical arrangement comprising: an array of photo-sensitive cells, each of substantially equal size, definable as d; and each cell providing an output related to the extent light impinges on the cell; means for directing a light spot to cells of said array, the spot size being definable as w, where d < w < 2d, whereby the light spot impinges upon at least two adjacent cells; and signal processing means connected to said cells to receive the outputs therefrom for generating a signal whose value is a function of the outputs of at least two adjacent cells.

2. An optical arrangement according to Claim 1 wherein the function is definable as

where i and j represent two adjacent cells and C, and Cj are the outputs of cells i and respectively.

3. An optical arrangement according to Claim 2 wherein said array comprises a linear array of a preselected number or cells.

4. An optical arrangement according to Claim 1 wherein said signal processing means include means for generating signals the value of each being a function of the outputs of different sets of adjacent cells and for selecting the value of the signal from a function when the value is within predetermined limits for the function.

5. An optical arrangement according to Claim 4 wherein the function is definable as

where i and j represent two adjacent cells and C and Ci are the outputs of cells i and respectively.

6. An optical arrangement according to Claim 5 wherein said array comprises a linear array of a preselected number of cells.

7. An optical arrangement for guiding a robot or the like along a path as a function of determination of the height of each incremental section of the path, the arrangement comprising: a source of light for directing a light beam to the path's incremental section, with light being reflected therefrom as a function of section height; an array of photosensitive cells in the path of the reflected light, the size of each cell being substantially equal and definable as d; optics means between said path and said array for controlling the size of the light spot impinging on the cells of said array to be w where d < w < 2d, whereby at any time the light spot impinges on at least two adjacent cells; and signal processor means responsive to the outputs of said cells for processing said outputs to provide a height indication signal as a function of the cells on which the light spot impinges.

8. An optical arrangement according to Claim 7 wherein said signal processing means includes means for generating functions for adjacent cells definable as i and j the function being

where C and Ci are the outputs of adjacent cells i and j, for selecting a value of one of said generated functions which falls within predetermined limits, and for providing a height indicating output signal which is related to the selected value of one of said functions.

9. An optical arrangement comprising: a beam of light directed substantially in-a selected direction; and optic means in the path of said beam for producing an image of the beam of a selected size, independent of the size of the beam impinging on said optic means.

10. An optical arrangement according to Claim 9 wherein said optic means includes an element with one side thereof defining a plurality of cylindrical convex lenses.

11. An optical arrangement substantially as hereinbefore described with reference to, and as shown in, the accompany drawings.