Mohan, William L. (Barrington, IL)
Willits, Samuel P. (Barrington, IL)

04/806371

02/15/1972

03/12/1969

Download PDF 3643068       PDF help

Click for automatic bibliography generation

Spartanics, Ltd. (Palatine, IL)

235/462.030

235/471, 235/494

G06C27/00; G06K7/10; G06K7/14; G06K19/06; G07G1/10; G06K7/14; G01N21/30; G06K19/06

235/61.12,61.11,61.115,61.9 340/146.3 250/219,202,203 88/14

3229100	Electronic servo system	January 1966	Grianias
3239674	Radiant energy receiving and detection systems	March 1966	Aroyan
3409760	Machine readable merchandise tag	November 1968	Hamisch
3413447	Information-bearing label and reading method and apparatus therefor	November 1968	La Mers
3414731	Package classification by tracking the path of a circular label and simultaneously scanning the information on the label	December 1968	Sperry
3418456	Encoded tag reader	December 1968	Hamisch

Wilbur, Maynard R.

Kilgore, Robert M.

We claim as our invention

1. Means for detecting and decoding a single channel of information consisting of a start code and a plurality of data bits, each bit comprising a pair of alternating contrast areas, the relative angular extent of each contrast area of the pair defining the binary status of the data bit, said single channel of information being arranged in annular form on a medium, comprising

2. Means for detecting and decoding a single channel of information in accord with claim 1 wherein said sensor means has a radial length to tangential width ratio of at least three to one.

3. Means for detecting and decoding a single channel of information arranged in annular form and having a discrete start code disposed on a medium, comprising

4. Means for detecting and decoding concentrically coded labels consisting of a fixed number of alternate light and dark annular bands phase modulated with respect to each other, comprising

5. A system for detecting information on a circularly coded label comprising

6. An identification system comprising

7. An identification system comprising

1. Field of the Invention

This invention relates, in general, to means for encoding and decoding labels such as used on produce or any other items to allow detection and recording of the items as they pass a checking station to provide continuous inventory and/or checkout control.

2. Description of the Prior Art

Inventory controls have long presented a problem and the checking out and recording of produce or other items has generally been done manually. For example, in a grocery store the clerk looks at the price on each item and records it on the cash register and moves the item into the group which have been previously recorded. Such controls are tedious, time consuming and subject to error and semiautomatic systems have been developed wherein each item to be checked out carried a coded ticket which could be torn from the item and inserted into the data handling system for the recording and accumulation of the price, inventory and other data. Such systems remove the human error which could occur in the previous manual systems wherein the checkout clerk might misread the price or might punch the wrong knobs on the cash register, for example.

However, tear-off tickets are subject to being accidentally removed and lost and require that each item be handled to remove the ticket and insert it into the data system.

Certain other automated systems such as readers require that the data be aligned and oriented before it can be entered into the data handling system.

SUMMARY OF THE INVENTION

The present invention relates to means of encoding and decoding labels such as would be used on produce, staples and any items sold in general merchandise stores, for example, for the purpose of supplying both the purchaser and the merchandiser with pertinent data relative to the item upon which it is placed. Although the invention is of particular applicability to stores, it is to be realized, of course, that the invention is also applicable to inventory control as, for example, in a factory wherein the receiving and disbursing of parts is maintained.

In particular, a coded label may be attached to articles of any shape which may in turn be intermingled so that the system is not limited to having items of the same size and the label may be decoded by a suitable hand probe decoder to provide data for the automatic tabulation of the item as to price, weight, item designation or any other required data needed in modern computer control business transactions.

Upon decoding the label the sensor probe and its associated electronics is capable of supplying data to an electronically operated cash register to show the purchaser and the operator the indication of the item cost and thus eliminate the human element in a mechanical tabulation of purchased items as now occurs in general merchandising.

The essence of the invention is that the coded labels used with the hand probe decoder is of such a design that specific orientation or positioning of the label relative to the probe is not required for the proper decoding of the label as is the general case with related devices.

The hand probe sensor may be an intensity measuring device utilizing the more usual quantities of optics relating to intensity and need not satisfy an equation of motion or boundary conditions as do devices utilizing diffraction techniques and requiring coherent illumination or self-luminous targets.

It is an object of this invention to provide means for encoding and decoding labels wherein the label may have a random orientation and displacement of its axis relative to the sensor probe.

Another object is to provide a device that does not require symmetrical placement of the label relative to the sensor.

Another object is to provide means for generating digital signals sequentially and/or parallel representative of data stored on a coded label.

Yet another object of the invention is to provide data to an electronic cash register and/or computer such that substantially all results previously obtained by manual actuation of the cash registers may be accomplished without error and very rapidly without requiring the cashier or clerk to read and manually punch information into the cash register.

Another object of the invention is to provide a coded label that can be permanently affixed to the item being purchased and can be read from any position by a scanner as it passes by a checkout point.

Other and further objects, features and advantages of the present invention will become apparent to those skilled in the art from the following detailed disclosure thereof and the drawings attached hereto and made a part hereof.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a perspective view of the random oriented decoder of this invention installed at a checkout counter of a supermarket;

FIG. 2 is a partially cut away view of the hand probe of this invention;

FIG. 3 is an exploded view of a scanning system of the invention;

FIG. 4 is an exploded view of a modification of the scanning system of the invention;

FIG. 5 is an exploded view of another modification of the scanning system of the invention;

FIG. 6 is a plan view of a label of the invention;

FIG. 7 is a plan view of another type label;

FIG. 8 is a plan view illustrating the relationship between the label and the hand probe;

FIG. 9 is a schematic view of a system for detecting and decoding information;

FIGS. 10A through 10L illustrate wave shapes in various portions of the system;

FIGS. 11A through 11G illustrate various wave forms obtained in the system of the invention;

FIG. 12 is a partially cut away view of a modified sensor of the invention;

FIG. 13 is an exploded view of a modified scanning system of the invention;

FIG. 14 is a schematic view of a system according to the invention;

FIG. 15 is a plan view of a label according to the invention illustrating various relationships;

FIGS. 16A through 16L illustrates wave forms in the invention;

FIG. 17 illustrates a label usable with the invention;

FIG. 18 illustrates another label usable with the invention;

FIG. 19 is a schematic view of a system according to the invention;

FIGS. 20A-20G illustrate wave forms in a system of the invention;

FIG. 21 is an exploded view of a modification of the invention;

FIGS. 22A and B illustrate a label and sensor orientation;

FIGS. 23A through D illustrate wave shapes appearing in a system of the invention;

FIG. 24 is a plan view of a label of the invention;

FIG. 25 is a partially cut away view of the hand probe of the invention;

FIG. 26 is an exploded view of a modification of the invention;

FIG. 27 is a schematic view of a system according to the invention;

FIG. 28 is a plan view of a label according to the invention;

FIG. 29 is a plan view of another label;

FIG. 30 is a block view of a system according to the invention;

FIGS. 31A through I illustrate various wave forms in the invention;

FIG. 32 illustrates a modification of the apparatus of the invention;

FIG. 33 illustrates a label of the invention; and

FIG. 34 illustrates a modification of FIG. 21.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

FIG. 1 is a perspective view of a pair of checkout counters 16 and 16', each of which have hand probes 11 and 11' connected to electrical cables 12 and 12' that are supported by supports 13 and 13' and which connect the hand probes to registers and computers 14 and 14' respectively. Customer monitoring stations 17 and 17' are mounted on the counters 16 and 16' so that the customer may monitor the information.

The package 19 is then positioned on the counter 16' under the probe 11' and has a label (not seen) which is being read by the probe 11'.

FIG. 2 is cut away view of a first form of the probe 11 and comprises a housing 22 which has an extension 23 with an operate button 24 and includes a sensor array (not shown) and a rotating optical wedge 26 which is driven by a motor mounted in the extension 23. A lens 27 focuses the image through the wedge 26 on the sensing array and is mounted above the label 21. The lower portion 28 of the probe 11 is transparent so that the operator can see the label and approximately align the probe over the label. The portion 28 might be of glass or a suitable plastic, for example.

FIGS. 6 and 7 illustrate respectively, different forms of coded labels, the one illustrated in FIG. 6 designated by 21 and having a plurality of pie segments 22, 23 and 24. The label 26 illustrated in FIG. 7 has a plurality of pie-shaped segments 27, 28 and 29. FIG. 8 illustrates the relationship between the label 21 and the end 31 of the probe 11.

The present invention allows lateral and angular misalignment of the probe 11 and the label 21 and still correctly detects information on the label. In FIG. 8, RH is the radius of the sensor probe see-through housing. RD is the radius of the coded label. RS is the radius of rotation of the sensor image and D is the displacement of the center of the scan from the center of the coded label.

The displacement requirements are that D must be less than RH minus RD where RH is greater than RD plus D and 2RS is greater than RD with the center of scan and the center of the sensor probe housing being coaxial.

The phase modulated label 21 illustrated in FIG. 6 comprises n number of binary bits plus an end of cycle bit designated as n+1. The end of cycle bit is 22 in FIG. 6. Each other binary bit comprises a light and dark contrast area with a total angular width being φ. The beginning of each bit segment is a light area and if the bit is a true (1), the light area will have an angular width (θ ₁ ) where θ ₁ = 3θ ₂ . If the bit is false (0), the light area (θ ₁ ) will have an angular width of 3θ ₁ =θ ₂ . It can be seen that with the randomly placed label 21 being scanned by a circular scanning sensor whose image is shown as (P) in FIG. 8 that the data generated by the scanning sensor will be frequency modulated, phase modulated train of pulses. The time interval per bit (Tb), assuming a constant sensor image velocity (Vp), will be a function of the effective radius of the scan (R _L ) relative to the center of the coded label. Where

R _L =√R _s ² +D ² -2DR _s Cosα

and

T _b =(R _L φ) / (V _p Cosβ)

where φ is the angle per bit in radians and the angle β is much less than 30°. If the radius of scan approaches the diameter of the coded label then the time per bit is

T _b =(R _L φ) / (V _p Cosβ)

The ratio of maximum time per bit to minimum time per bit is

T _s max./ T _S min.=(R _S +D) / (R _S -D)

The limitation of the angular requirement per bit is a function of how large a coded label is desired and the resolution of the scanning sensor. For the example shown in FIG. 8, a one inch diameter label containing fifteen bits is shown as a drawing enlarged six times. The see-through probe has a diameter 15 percent larger than the coded label's diameter and the diameter of the circular scan is 20 percent less than the diameter of the coded label.

There are many ways for nutating an image to produce a circular scan and FIG. 3 illustrates one such method. It is to be realized that the housing of the sensor and the mechanical supporting portions have been removed for purposes of clarity. A lens 36 is mounted in a rotatable housing 37 which is supported in the probe 11. The lens 36 is mounted so that it is offset from the center axis of the housing 37. The housing 37 is driven by a motor 38 which has output shaft 39 that carries a driving gear 41 that engages teeth formed in the edge of the housing 37. A sensor array 42 receives an image through the lens 36 from the label 21 as shown by the dotted line on the label 21. The electrical data thus generated in the sensor array 42 by scanning the light and dark contrast areas of the label are amplified in an operational amplifier 43 and supplied to output terminal 44. FIG. 4 illustrates another way of effecting a circular scan. The output gear 41 of motor 38 drives a rotatable housing 46 mounted in the probe 11 which supports an optical wedge 47 which changes the direction of the impinging light rays to cause the image of the sensor 42 to traverse a circular path as shown by dotted line on the label 21. The field lens 48 collimates the image of the sensor 42 adequately to allow for a change in depth of the front focal position of the coated label when it is not placed upon a flat surface. The electrical output data from the sensor array 42 is supplied to output terminal 44 through operational amplifier 43. A further lens 49 may also be included in the system if desired.

FIG. 5 illustrates another method of producing a circular scan wherein the motor 38 through the gear 41 rotatably drives an opaque disk 51 which has a small opening 52. The distance of the opening 52 from the optical axis 53 and the magnification factor of the objective lens 54 will determine the radius of scan. The disk 51 is caused to rotate by the motor 38 and allows the output from the label 21 to pass through the disk aperture only along its effective path as shown by the dotted line to the sensor array 42. The collecting lens 56 focuses upon the sensor array 42.

The operation of the decoder relative to the scanners illustrated in FIGS. 3, 4 and 5 is shown schematically in FIG. 9 with relevant wave forms illustrated in FIG. 10. The label of FIG. 6 is used in this embodiment.

The sensor array is chosen so that an effective radial length to tangential width is greater than three to one for purposes of good signal-to-noise ratio (optical-electrical) and the electrical signal corresponding to the optical contrast gradient from bit to bit will be as shown in FIG. 10B as a sensor scans the label. FIG. 10A shows the wave form B in linear form. The start of the code series is at T ₀ and the end is at bit n+1. The electrical data is fed from the operational amplifier 43 to an amplifier 60 which includes an amplifier 61 and a logic gate 62. The logic gate 62 determines that the data used is above a certain minimum as defined by a reference voltage applied to terminal 63. The output of the amplifier 60 is supplied to a differentiator 64 which produces an output wave form shown in FIG. 10C. The positive going spikes are used to trigger the "true" one-shot multivibrator 66 of stage 67. The negative going spikes are used to trigger the "false" one-shot multivibrator 68 of unit 67. These gates are shown in FIGS. 10D and 10E, respectively, and are used for multiple purposes. The time duration of the "one-shot" gate is very short compared to the minimum interval generated by a given bit and is on the order of 5 percent or less. The leading edge of wave form 10D is used to generate the start of the true gate in the true-false computer gate logic 69. While the leading edge of wave form 10E is used to generate the end of the "true" gate of element 69.

The true gate is wave form 10F while its compliment is the false gate shown as 10G. Assuming that the data brightness is valid for the given portion of the cycle that is defined by 10F and 10B then the gates 10F and 10G will be used to operate bit analogue computer 71.

As stated, the present type of scanning produces a frequency modulated, phase modulated signal train. So as not to have ambiguities in the bit data it is necessary to define a bit's truth whether "true" or "false" strictly on a cycle-per-cycle basis; and thus the phase modulation method is used. Since the time per bit is a variant in the case disclosed in FIG. 8 by as much as 2 to 1, then each bit is defined relative to its own complete cycle.

If the bit analogue computer 71 has a constant integration which is a fixed rate of change of voltage with respect to time as defined by input reference voltages plus reference at terminal 72 and negative reference at terminal 73 being equal and opposite in polarity, then the polarity of the output at terminal 74 is determined by the ratio of the gate time that connects the input first to the "minus reference" through resistor R5 and transistor Q4 via gate 10F and then to the "plus reference" by resistor R4 and transistor Q3 via the gate 10G.

The operation of the analogue computer 71 begins at T ₀ with the output of M3 being 0 volts. With the appearance of gate 10D the true gate 10F starts and electrically connects the input of the integrator to the negative reference voltage which has a magnitude high enough to insure that for any given minimum bit time the output will be greater than 3 times the trip level of E1 or E2. False gage 10G is now connected to the integrator input after a slight delay period, defined by gate 10E and now causes the integrator to integrate this reverse and exactly opposite magnitude of current to reduce the output voltage of M3. At the end of false gate 10G in combination with gate 10D, the output polarity of M3 is sensed. If the true gate is longer than the false gate, as previously defined by the ratio of θ1 to θ2, FIG. 10G, the bit information is true or 1, as the polarity of the output of M3 is above E1 of FIG. 10H. This train of true-false bits, having a polarity of either plus or minus for 1 or 0, binary data, is sequentially fed to the computer. Portions of it can be gated out by "N" bit ring counter 76 for local display.

At the end of the "n" bit of data, the n+1 bit operates the integrator as before. The difference being that the false gate generated by the long dark area of the coded label generates a much longer false gate so that the integrator reaches a negative value far greater than any generated by a false gate where the time per bit is at a maximum. This causes the output of the integrator to exceed the value of the negative reference E3 and in doing so trips the recycle bit generator 78 which restores the integrator to zero and resets the ring counter 76 in addition to supplying the information to both the remote and local computer.

The gate 10D is used to read out the integrator's polarity and at the end of this gate is used to reset the integrator to zero via the transistor Q5, diode D5, diode D6 and capacitor C2.

The operation of the decoder relative to the methods illustrated in FIGS. 3, 4 and 5 with a label such as shown in FIG. 7 is illustrated graphically in FIG. 11. Assuming that the label is randomly placed under the sensor and the dotted lines on FIG. 7 represent the circular scan track of the sensor, then the output of the sensor will be as shown in FIG. 11B comprising a complex frequency modulated, phase modulated train. This data can be handled by the same system illustrated in FIG. 9 and will require few circuit changes.

The positive going spikes of the differentiated signal out of the differentiator 64 are used to generate gate 11D and the negative going spikes to generate logic gates 11E. The leading edge of the first positive going pulse 11D starts the true gate 11F while the leading edge of the next 11D pulse stops the true gate. Pulse gate 11G is the compliment of this. Gate 11G is used to insure that the right sequence of 11D gates are used such that the 11E gate always starts after the 11D gate.

The analogue bit data is generated as before and the readout of M3 is done by gates 11D with gate 11E inhibiting the wrong gate 11D.

This system has an advantage in that the the basic contrast area can have the same angular requirement for ease of manufacturing and the light or dark area is marked by its opposite contrast to give the body of the label the required code.

The systems discussed so far have simple sensor mechanisms and use simple label designs. The systems to be described now are capable of larger bit density per label and may be adapted to parallel bit multiple sensor readout.

FIG. 12 illustrates a sensor which has an L-shaped housing 80 with an extension 81 for enclosing a drive motor. Extension 82 is transparent so that a label 83 on a package 84 may be observed by the operator. Image-positioning means 86 are provided to optically align the rotatable prism 87 which is driven by the motor within the housing 81 so that the principal axis coincides with the center of the coded label. The displacement distance D is illustrated in FIG. 15, for example. In order to generate parallel sequential data based upon circular scanning it is necessary to effectively rotate the image of the detector array (an array being one or more detectors extending radially from the axis of rotation) about an axis which is essentially coaxial to the center of the coded label. This requires that prior to decoding the bit data stored in concentric areas about the center of the label, it is first necessary to develop data to position the optical center of scan to the center of the randomly placed label.

FIG. 13 illustrates one embodiment of this invention for automatically positioning the optical axis of the scanner to that of the label. The detector comprises an array designated as 42a, b, c, d and e upon which the image is focused by the lens. The innermost sensor 42e which is at a finite radius from the optical axis 100 will be used to automatically center the optical axis to that of the center of the label 83.

A double dove prism 89 is mounted in a rotatable housing 87 which is driven by the motor 138 through the gear 141. To most efficiently use a double dove prism it is desirable to transmit essentially parallel rays through it. The sensor array 42 located at the principal back focal plane of lens 99 will have its image collimated by said lens to produce the required parallel ray bundle necessary for transmitting through prism 89. Upon leaving prism 89 the rays are then refocused upon label 83 by lens 101 after first impinging upon image translating mirror 91 of the centering mechanism 86. The direction of translation of the mirror is controlled by torque motors 93 and 94 which are mounted on suitable axes connected to the brackets 92, 96 and 97 to move the mirror to center the image to the gimbal system. The output of the detector 42e controls the motors 93 and 94 to center the image.

A label illuminating source 98 provides lambertian illumination which is preferable for many reasons, the most important reason being that specular illumination can cause false data due to highlight on the label. Further, since the tungsten lamp has most of its energy in the infrared region, a suitable IR band pass filter can be placed in front of the detector to eliminate data due to local ambient illumination which is usually in the visible spectrum when generated by fluorescent lamps.

Motor 138 is a polarized synchronous type and is used to rotate the prism 89 in the housing 87 about the optical axis of the prism. For every 360° rotation of the prism, the image will be rotated 720°. Thus, a 2 to 1 stepdown gear 141 is used to synchronize the prism position relative to its u- v axis and to the 60 cycle excitation applied to the motor 138.

Utilizing the detector 42e for centering the optical axis to that of the label of the types illustrated in FIGS. 17 and 18, the geometry of the label centering spot to the innermost sensing element 42e is defined in FIG. 15.

The data generated for centering by this innermost detector 42e is illustrated in FIG. 16B based upon the conditions illustrated in FIG. 15. For any given position of the detector axis for rotation to the center spot of the label, a corresponding phase dependent contrast gradient will be sensed. Using the 60 cycle line as a reference to generate appropriate quadrature gates 16C, D, E, F, G and H as shown in FIG. 14, the 60 cycle reference line is supplied to a polar demodulator 106 to generate quadrature gates shown in FIGS. 16C, D, E, F, G and H and the direction in which the center spot is off axis to the detector axis can be detected in demodulator 107 of FIG. 14. For each axis of correction there is a requirement for a plus or minus correction and the error amplifier and filter 110 receives the output of the polar demodulator 107 and produces two outputs that are fed to the motors 94 and 93 through the amplifiers 111 and 112 through terminals 113 and 114 respectively. The logic gate 116 also receives the centering information and produces an output at terminal 117 which indicates when the image is centered.

Using the centering technique described above, decoding of the "phase marked" n bit sequential label 118 of FIG. 17 is shown in FIG. 19 and associated FIG. 20 which illustrates waveforms present in various portions of FIG. 19. As shown and described with reference to FIG. 14, the innermost sensor 42e is used for centering the optical axis to the center of the label so that each sensor has a different radial position and will effectively scan successively larger concentric circles. For label 118 in FIG. 17, two sensors are required in addition to the centering sensor 42e; one, 42a, for generating the clock data 20A and the other, 42b, for generating the bit (Mark) data 20B relative to the phase of the clock data 20B. FIG. 19 shows schematically the simplicity of the decoding where each of the two additional sensors has its own preamplifier, 119 and 120. Further amplification and brightness level validation are provided in gate 300, differential amplifier 301, clock gate and true-false gate in 302, in a manner similar to that described above in conjunction with FIGS. 9 and 10.

The truth logic for a bit is straightforward requiring the mark sensor 42b to generate a gate relative to the light or dark area of the clock. If the bright contrast area signal from sensor 42b appearing at lead 122 in FIG. 20 is generated while the bright contrast area is generating clock signal A as shown in FIG. 19 then the bit is true; if it appears in the dark area then it is false. If there is no mark in one clock cycle then this lack of a mark will be used to recycle the system. The recycle gate generator 123 in FIG. 19 provides this logic component. As before, ring counter 124 is fed by the clock gate to generate gating pulses for local readout of selected bits if required. Brightness logic gate 126 insures that the bits generated are above a certain minimum contrast level for data validating. A gated bit local display readout 127 is provided and the true-false bit generator 125 provides the output to the computer.

The label 128 illustrated in FIG. 18 is another type of coded label requiring automatic centering of its axis to the center of scan, this type being an n bit, n word sequentially coded label, the word length being the number of bits radially from the center and the number of words being the number of clock cycles per 360°.

For every concentric ring containing data relative to the system, whether for centering or bit readout, there will be a requirement for a sensor array having the same number of sensors optically matched to their corresponding circle of rotation.

The method of handling the data for the label shown in FIG. 18 is essentially the same as that discussed with respect to the label of FIG. 17 with the changes being that for each sensor there is a corresponding preamplifier, brightness level amplifier, differential amplifier and true-false gate generator.

The label illustrated in FIG. 18 may be considered an "n" word, BCD coded label, then the four outermost bands 129, 130, 131 and 132, respectively, are used to define the weight of the word and the fifth band in from the outer edge 133 may be used to reset the system. The sixth band 134 is the clock band: the truth table being that for any given bit in a given band that has a bright contrast area relative to the clock's bright area the bit is true (1). For any given bit in a given band that has a dark contrast area relative to the bright contrast area of the clock the bit is false (0). The fifth concentric ring of data 133 is shown in FIG. 18 as all false until it is time to recycle where a true bit is shown as is designated at 133a.

Many variations of labels may be used and the ones illustrated are, for example, types that may be used with a random orientation of circular coded label of this invention.

FIG. 21 illustrates another modification of the invention wherein a single detector 142 scans the label through a pair of slotted members 156 and 157 which form an orifice that scans the label 147. The slotted members 156 and 157 are driven respectively by motors 154 and 153 which cause them to oscillate suitably so that the orifice formed scans the label 147. Lenses 151 and 152 are mounted in the optical path and a cylindrical lens 149 mounted in a rotatable support 148 is driven by motor 143 through its output shaft 144 and gear 146. The cylindrical lens 149 scans a rectangular-shaped portion on the label 147 as the scanning traverses about the label 147. The sensor 142 is used to decode the bit data stored on the label as well as to center the sensor relative to the label.

It is of great importance that the effective aperture which, has essentially the same dimensional ratio of width-to-length, be optically changed by cylindrical lens 149 into a radial line sensor. In this case, an optical wedge 148 so configured so as to give the proper displacement of the optical axis for scanning the label is used.

The image rotating means have been described with other embodiments and will not be discussed in detail with reference to this embodiment. The cylindrical lens 149 is so placed upon the rotating optical wedge 148 so as to convert the square image of the aperture to a radial line image whose length-to-width ratio is determined by the power of the cylindrical lens. This ratio, as discussed above with respect to the sensor of FIG. 9, is of prime importance for the elimination of false data.

FIG. 24 illustrates a label 160 for use with the system of FIG. 21. The geometry of the displacement of the label relative to the probe is illustrated in FIGS. 22A and 22B. When the label is randomly placed under the sensor probe, the data generated by a single sensor having the requirements as set forth in FIG. 22, is a phase modulated, amplitude modulated train of pulses as shown graphically in FIG. 23C. FIG. 23D shows the same data after suitable filtering. Comparing this wave train to that out of the squaring amplifier 95 in FIG. 14, it will be seen that the two signal trains are essentially identical in nature. It is only necessary to handle this data as in the system of FIG. 14, as shown in FIG. 19, for automatic positioning of the effective aperture of the probe in FIG. 21 to generate a circular scan whose center of rotation is the same as the label under the probe.

FIG. 23B illustrates the data generated by the scanning aperture once the system has been centered. This data is the same "phase modulated" train of bit data as previously discussed in the system having a single detector with the exception that the frequency modulation characteristic of the wave train is missing. Elimination of this undesirable characteristic gives a greater ability for high bit density per circular scan.

Since this is an automatic centering device of the axis of scan to the axis of the coded label, it is necessary in this embodiment to correlate the u-v axis of scan to the label as the image is rotated. Thus, the drive motor of the optical wedge is driven by the polarized synchronous motor 143 through a 1 to 1 gear ratio. The systems previously disclosed are based upon a sensor probe having built into it the electro-optical and mechanical means necessary for proper image positioning and scanning of the randomly placed label under its "see-through" housing. It can be advantageous to perform these operations at a remote location in order to eliminate the possibility of undetected damage to the probe's scanning mechanism caused by rough handling.

A means for accomplishing this is shown in FIG. 25. The hand probe 11 has a see-through portion 177 which has a bottom edge 165 that rests over the label 171. The housing 170 includes a lens 172 which focuses the image of the label on an image tube 173 which has a face plate 174. Image tube scanning beam deflection means 176 and the necessary wiring go out the cable 12 to the remote decoder. The housing for the probe can be constructed of a tough resilient plastic and the image tube can be ruggedized and of a high-sensitivity type.

Decoding the latent image on the face plate of the image tube can be accomplished in a number of ways.

The image of the label can be transmitted as in conventional close-loop TV systems and be reconstructed on a suitable kinescope. The methods described for decoding the label in the previously mentioned systems with their associated coded labels can also be applied to the new remote image of the label and at any scale factor desired.

For best results, the scanning period of the optics should be at a lower frequency and asynchronous to the raster scan of the image tube.

Another method for remotely decoding those labels that are sequentially coded and require circular scanning is to generate a circular sweep for producing a circular scan directly on the latent image on image tube 173. The geometric requirement of this circular scan will be similar to the equivalent sensor scan of the prior systems. The video data out of the image tube will also be similar to those systems and can be decoded as explained before.

For those labels requiring multiple sensor decoding as in the prior multiple sensor systems, the sweep generator necessary for producing a circular scan directly on the latent image on the image tube will be more complicated.

Since, in general, the image tube has a single electron gun for generating a scanning beam and the position of this scanning beam is controlled by the image tube deflection means, it becomes necessary to generate sequential circular scans whose number and radii are governed by the previously discussed geometry relative to that label.

This system of multidiameter circular sweeps will require a sequential high-speed gate generator so that for each different diameter of sweep there is an associated time coincident gate to properly channel the video data to its associated electronics.

FIG. 30 illustrates the circuitry for generating a circular sweep.

Generating a circular sweep on cathode ray tubes is an old art as is the use of DC on deflection coils for translation of the beam. For the generation of the multiple diameter circular sweeps for the image tube 180 in FIG. 30, the sine-cosine generator 181 develops as many different amplitude of the input reference f ₁ into a sine and cosine relation as there is a need for different radii of scan. For each radius of a given diameter the high-speed electronic switches 182 and 183 sequentially gate in the required amplitude of sine-cosine voltage E ₁ , E ₂ and E ₃ representing the beam deflection coordinates, (x)=E sine ω ₁ t and (y)=E Cos ω ₁ t. Thus, the scanning beam of tube 180 is sequentially moving in discrete steps from a small diameter to a medium diameter to a large diameter at a very rapid rate while being rotated at a much slower rate.

With each radial position and tangential position of the beam there is an associated video signal representative of the contrast gradient at that time and place of scan. Consequently, the scanning beam video signal is a complex wave form as shown in FIG. 3II, with time expanded per Δto

In order to separate this complex signal into the required components, it is necessary to gate out the video data representing a given radius at the time the beam is at that position. Thus, for each gate generated by n ring counter 184 in FIG. 30 for gating in a particular value of sine and cosine of the reference f ₁ , the same gate is used to gate out the various video data representing those radii of scan from oscillator 186 in FIG. 30, representing the radii of scan. FIG. 31 shows similarity of this data 31B, C and D as compared to the data in FIGS. 16 and 20, respectively.

Once this data has been smoothed by filters 187, 188 and 189 in FIG. 30, it can be applied directly to the appropriate systems illustrated in the prior drawings for automatic centering and bit decoding.

As mentioned before, the high frequency oscillator 186 should have a frequency asynchronous to the circular scan frequency f ₁ and a repetition rate for generating sequential gates in the ring counter 184 with the following requirements:

f ₂ should be greater than 10(n+1) af ₁ where

(n)= number of bits per 360° scan

(a)= number of concentric scans required

(f ₁ )= reference frequency, 60 cycles

The labels in systems described so far have utilized circular scanning and circular coded labels. It may at times be advantageous to use linear scanning for labels such as illustrated in FIGS. 28 and 29. For this embodiment the sensor may consist of that shown in FIG. 25.

In order to decode a random oriented label as shown in FIG. 29 without resorting to complex cross correlation techniques, it is necessary to place certain requirements on the random displacement of the "see-through" housing of the sensor probe relative to the coded label outside centering ring 190 in FIG. 29.

Using the displacement requirements of prior systems applied to label 191 where:

C= center of label

S= effective center of centering scan

R _s = mean radius of centering scan

R ₁ = inside diameter of centering ring on label

R ₂ = outside diameter of centering ring on label

L= effective radial length of centering scan beam

D= displacement of center of scan to center of label = L/2 where R ₁ is much greater than L; R ₂ -R ₁ is much greater than L; and R _s =R ₁ +L/2 and center of the image tube is coaxial to the center of the "see-through" housing.

Since the image tube scanning beam is also multiplexed in this embodiment and the beam, when sharply focused, has little physical dimension, then it becomes necessary to give the beam an effective radial length L in order to generate representative data of the contrast area involved.

As previously stated, the deflection x,y coordinates of the beam are x=E sine ω ₁ t and y=E sine ω ₁ t, where the amplitude of E determines the radius of scan. If we add to this another sinusoidal voltage of amplitude Δ E, whose frequency f ₃ is much greater than f ₂ and multiply it by the sine, cosine function of E, we will have a radial scanning beam of effective length Δ E with a mean radius of E. Instantaneous voltage representing the beam (x)(y) coordinates are now X=(E+ΔE Sine ω ₂ t) Sine ω ₁ t and 4=(E+ΔE Sine ω ₂ t) Cos ω ₁ t. WITH REFERENCE to FIG. 30 the voltages E ₁ and E ₁ ' from sine-cosine generator 181 represent the mean (x)(y) coordinates of the centering scan. To get an effective radial length (L) of rotating scan beam going to switch SW (1) 182 and switch SW (2) 183 is replaced with a voltage Δ E from an additional high frequency oscillator f ₃ , summed with E of reference f ₁ and then to a sine-cosine multiplier before going to the respective switches 182 and 183 for generating the new centering scan beam.

Once the concentrically coded label 191 of FIG. 29 has been centered, it is only necessary to generate an additional bit decoding beam to linearly scan either the X or y axis to decode the label. The type of coding illustrated in FIG. 29 is phase modulated where the first bit cycle is from the beginning of the second light contrast ring to the beginning of the third light ring and so forth. The first light contrast ring 190 is used for centering. The bit data generated will have a mirror image whenever the bit scanning beam crosses the center of the label. This can be used for validation if the data is properly stored and compared to its mirror image.

For those labels, such as shown in FIG. 28, where the bit data is stored in raster form and n line of n bits are contained in a given raster field, it becomes necessary to determine this field's orientation relative to the axis of scan, whether electro-optical or image tube scanning is used for decoding. FIG. 26 illustrates the apparatus and method for centering a label and for determining the orientation of the raster. The method used is with the centering ring image 193 of label 192 as shown in the position of this band of light contrast area is controlled about the y axis by the two cells 196 and 197 which receive signal through the lenses 198, the rotating means 199, the lens 201 and the centering mechanism 202 which has centering motors 203 and 204. As shown in FIG. 27, the output of cells 196 and 197 are first amplified in amplifiers 206 and 207, respectively, and are then differentially compared in amplifier 208 and any resulting error is used to achieve a balance by servo-amplifier 209 which operates torque motor 204.

The position of band 193 is controlled in a like manner about the x axis by sensors 211 and 212 shown in FIG. 26 through amplifiers 213 and 214, 216 and servoamplifier 217.

This centering is accomplished while the prism 218 is not in rotation and in its home position as determined by cam 219 in FIG. 26, switch 221 and image rotation logic unit 222 in FIG. 27. Once the label has been centered and both x and y axis servos are at null, as detected by null logic 223 in FIG. 27, the image is allowed to rotate as elements 222 and 223 dictate. Amplifier 226 determines when the absolute brightness of detectors 211 and 212 are twice the brightness of detectors 196 and 197. This, in combination with the logic element 222, knowing the label is centered, will then operate relay 228 to immediately stop the image rotating means through contacts 229 that are connected to terminals in circuit with the driving motor 231 which drives the prism 218 through the gear 232 and housing 233. Thus, the label image having been randomly placed under the "see-through" sensing probe will be ready for raster scanning and decoding.

The length L of cell A+B comprising the detectors 236 and 237 should be equal in length to the radial width of the centering ring 193. The area of the orientation mark 194 should be approximately equivalent to the area of a single cell and it should be centered radially in the light area of the centering ring with a radial length of one-half of the radial width of the centering ring.

FIG. 32 illustrates an embodiment in which the detector 240 is nutated by nutating mechanism 241 through a linkage 242 and is centered by a centering mechanism 243 through a linkage 244 relative to a label 246 rather than the systems disclosed wherein the image is rotated by rotating a wedge, or other structure. Devices for nutating are well known in the electronic art and are not described in detail herein. The output of the detector 240 would be processed in the same manner as that disclosed with reference to the other embodiments.

FIG. 33 illustrates a label 230 which has printed on it the value of the item so that the label may be read visually if desired.

FIG. 34 illustrates a modification of the image rotation portion of the system of FIG. 21 in which the housing 148 includes an opaque aperture 231 formed with a slot 232 which rotates with the cylindrical lens 149. Lens 151 of FIG. 21 may be a spherical lens and the combination of 149 and 151 focuses the aperture formed by members 156 and 157 on the label in the direction normal to the displacement between members 156 and 157. The cylindrical lens 149 having power in only one direction, allows the image to be focused well in one direction while being defocused at right angles.

Opaque aperture plate 231 is mounted with lens 149 and 151 and rotates with them in housing 148. The direction of the long axis of aperture 232 is parallel to the displacement direction of members 156 and 157. This provides good depth of focus at the label in a direction normal to the displacement while not reducing the light gathering power as a round aperture would.

It is seen that this invention provides means for sensing and detecting the information on randomly orientated labels and although it has been described with respect to preferred embodiments it is not to be so limited as changes and modifications may be made therein which are within the full intended scope as defined by the appended claims.