DETAILED DESCRIPTION OF THE INVENTION

[0038] A significant problem in reading decodable indicia such as bar codes and other types of symbols that can be interrogated optically is reading such symbols under widely varying lighting conditions. In practice, the lighting conditions for a reading can vary from fully acceptable lighting conditions, such as examining a parcel bearing a bar code under ambient sunlight at high noon on a clear day, to minimal lighting conditions, such as may be found in examining an unwieldy bar-coded part or parcel on the bottom shelf of a poorly lit warehouse at night. Often, it is impractical or impossible to control the lighting conditions. Nevertheless, it may be imperative to correctly read and decode one or more symbols to identify a particular object bearing the symbol.

[0039] The problem can also be understood in technical terms as the requirement to maintain suitably high levels of accuracy and precision in converting the analog signal of the light reflected from the symbol into digital data. Significant accuracy errors in performing the conversion occur when the minimum unit of expression, represented as the least-significant bit, is larger than necessary. Further errors in precision, in the form of reduced numbers of bits used to represent the value, occur when the range spanned by the converter is appreciably larger than the maximum magnitude that is to be converted, or is smaller than the maximum magnitude to be converted. In some embodiments contemplated herein, the digital data is evaluated so as to provide a suitable identification of the object bearing the symbol, or to provide an attribute associated with the object, such as a model, a price, an expiration date, or any other information about the identified object. In particular, one would like to have the same accuracy and precision from measurement to measurement independent of the lighting conditions. One would further like to have accuracy and precision that are uniform over the operating range of the measurement, without the necessity to pre- or post process signals and/or output data simply to obtain suitable accuracy and resolution. Those of ordinary skill in the data conversion arts will recognize that many other embodiments and applications of the principles of the invention disclosed herein are possible.

[0040] In general, analog phenomena do not express themselves in predetermined magnitudes, but rather as values in a continuum. By comparison, digital phenomena are defined in discreet magnitudes, (e.g., binary digits or “bits,” expressed as one of 1 and 0) commonly represented by two voltages, such as “HI” and “LOW,” or “on” and “off.” Converting an analog signal to digital form requires that a value having a magnitude that might be expressed as an integer, a fraction, or even an irrational number, be expressed in a form where the units of expression are limited to powers of 2, including positive and negative integer powers.

[0041] One approach that can provide sufficient accuracy and precision is to use an A/D converter that provides more range and precision than needed, and to select and scale the data it provides. However, such an approach has the drawbacks of the increased cost of an A/D converter having higher resolution (e.g., providing 10 bits when only 8 bits of resolution are needed), as well as increased computational issues in the analysis of data having different ranges and different precision as expressed in bits. For example, a bar code reading taken under bright light would produce a signal that has more digits of precision than a different reading of the same bar code taken under dim light, merely because the light intensity falling on the bar code changed. In many instances, such as for reading bar codes or the like, the measurements necessary to decode the encoded information needs to be of sufficient accuracy and precision to deduce the encoded information from the recovered signal, but need not be an absolute measurement. In other embodiments that require an absolute measurement, such as determining the angular position of the butterfly valve of Wright, absolute values can be obtained by a simple post-processing calculation to scale the final results. In one embodiment, the post-processing involves the multiplication of each digital data value by the inverse of the value of a rational number multiplied by a characteristic voltage, which value is discussed at greater length hereinafter.

[0042] A solution that optimizes both the accuracy and the precision involves scaling the span of the analog full scale representation to exactly match (or very slightly exceed) the dynamic range of the signal to be measured. In the present invention, this is accomplished by applying first and second reference voltages, V_ref1 and V_ref2, having a voltage difference V_ref1−V_ref2, to the power inputs of an A/D converter that produce a voltage range, expressed as the absolute value of V_ref1−V_ref2, that is just large enough to accommodate the dynamic range of the analog data signal to be converted, which range need not be expressed as a power of 2, as a power of 10, or as any particular multiple or fraction of a specific voltage. Because of the nature of A/D converters, and of D/A converters, it is convenient, but not absolutely necessary, that the applied voltage difference V_ref1−V_ref2 be expressible as a rational number multiplied by a voltage value.

[0043] Heretofore, scaling the span of an A/D converter to an arbitrary value that changes with time and/or with changes in the environment has not been an easy or a convenient matter. The problem of performing such scaling relates to either or both of a circuit that can amplify analog signals using a varying arbitrary constant, and to an A/D converter that can scale its output to match arbitrary limits. A solution for these problems is disclosed herein, and embodiments of the solution, in the form of circuitry, are described.

[0044] Turning to FIG. 1, a graph 100 depicts a digital representation of two analog signals 102, 104 that vary in space along a straight line. In other embodiments, the signals could equally well vary with time. The dimension 110 along which the analog signals vary may be expressed in units of measure, or may be expressed, as shown here in channel numbers, corresponding to the ordinal number of a measurement in a sequence of measurements. Here the ordinal number varies from zero (0) to a maximum channel number C, where C is a positive integer. In the embodiment described, the analog signals 102, 104 represent observations of reflections of light from a black and white pattern, such as a bar code, by each pixel of a sensor such as a CCD sensor, for example the TCD1304AP available from Toshiba of Japan, having a plurality of pixels, for example 3648 pixels.

[0045] Signal 102 represents an observation under a low light condition. Signal 104 represents an observation under stronger illumination. As indicated on the vertical axis 120 expressed in units of voltage, the CCD sensor provides a higher voltage as the intensity of the illumination is reduced. A maximum voltage, represented by V₃ on axis 120, which in one embodiment corresponds to 2.7 volts, is observed under conditions of minimum reflected illumination reaching the CCD sensor, such as when a pixel of the CCD sensor is observing a poorly reflective surface, such as a black portion of a bar code. A minimum voltage, represented as V₂ on axis 120 for signal 102, and as V₁ on axis 120 for signal 104, occurs when a pixel of the CCD sensor is observing a highly reflective surface, such as a white portion of a bar code. In this embodiment, under low illumination the value of V₂ may be 1.7 volts, while under stronger illumination, the value of V₁ may be 0.9 volts. Vertical axis 130 indicates the relationship of the observed light intensity to the voltages expressed on axis 120 and to the signals 102, 104. In other embodiments, the same kinds of observations can be observed for decodable indicia such as symbols having color, symbols having gray scale, symbols representing biometric information such as fingerprints, retinal patterns, or physical features such as facial features, and other types of symbols that provide an optical signal by using a suitable sensor that is sensitive to color, gray scale, or other representations of absolute or relative reflected illumination.

[0046] FIG. 2 illustrates an exemplary embodiment of a circuit 200 that provides optimal accuracy and precision in measuring and converting analog signals observed from decodable indicia such as bar codes. The circuit 200 is powered by any convenient electrical power supply, such as a battery or a power supply connected to a wall socket. Power supplies are well known and are not shown in FIG. 2. The circuit 200 includes an A/D converter 210, such as the ADC 1173 from National Semiconductor that receives control signals, including clock signal CK from the microprocessor 215, as indicated by the line labeled CK. In one embodiment, the microprocessor 215 is a programmable system on a chip such as the CV8C26233-24PVI PSoC™ available from Cypress Semiconductor of Seattle, Wash. Alternatively, the microprocessor 215 can be any one of several provided by Motorola, Texas Instruments, and other manufacturers.

[0047] The microprocessor 215 also controls two reference voltage sources, which in one embodiment are incorporated into the microprocessor 215 integrated circuit chip. For example, the reference voltage sources may be individual D/A converters that are each driven by a respective digital signal under the control of the microprocessor 215. A first reference voltage source provides a first reference voltage V_Ref1 that is applied to a first reference voltage input terminal of A/D converter 210, as indicated by the line labeled VRef1. A second reference voltage source provides a second reference voltage V_Ref2 that is applied to a second reference voltage input terminal of A/D converter 210, as indicated by the line labeled VRef2. The microprocessor 215 can set each reference voltage independently of the value of the other. By way of example, using M-bit D/As where M=8, the microprocessor 215 can set the first reference voltage V_ref1 to be any of 2^M=2⁸=256 values in the range of 0 to 255 times a characteristic voltage V, and the microprocessor 215 can independently set the second reference voltage V_ref2 to be any of 256 values in the range of 0 to 255 times the same characteristic voltage V. Therefore, the difference between V_ref1 and V_ref2 will be an integer times the characteristic voltage V. In one embodiment, the characteristic voltage V is 1/255 times a full scale voltage (e.g., V_fs) for the D/A converter, and the difference V_ref1−V_ref2 equals an integer/255 times V_fs, or a rational number multiplied by a voltage value, as described above. Those of ordinary skill in the digital data arts will recognize that, in the example given above, the fineness of the resolution of the reference signals is defined by the number of bits available in the D/A converter, which need not be identical to the number of bits provided by the A/D converter, but logically should be no fewer in order to accomplish the intended ends of the present invention. The D/A convert can have more bits of precision than the A/D converter, in which case the span of the A/D converter can be made arbitrarily close to, but greater than, the dynamic range of the signals to be measured. Also, as mention above, the inverse of the characteristic voltage multiplied by the rational number, or equivalently, the inverse of the difference in voltages between the two reference voltages, can be conveniently used to scale relative digital data to absolute digital data, as long as an absolute magnitude corresponding to one digital data measurement is know, for example by measuring a calibrated input signal.

[0048] The A/D converter 210 receives analog input signals at an analog input terminal 212. In one embodiment, the analog input signal comprises one or more signals from respective pixels of a sensor 220, such as a CCD sensor, that observes illuminated objects, such as bar codes under ambient or artificial illumination. In one embodiment, the CCD sensor is a TCD1304AP from Toshiba of Japan. Alternatively, the sensor 220 is an ILX526A manufactured by Sony of Japan. The sensor 220 may provide a signal directly to the A/D converter 210, or the signal may be provided by way of an optional switching circuit (not shown), such as a multiplexer (hereinafter “mux”) of conventional type known in the art under the control of the microprocessor 215, with or without a sample-and-hold circuit of conventional type (also not shown). The analog signal may be controlled temporally by a clocking signal, such as clocking signal CK, from the microprocessor 215.

[0049] In another embodiment, the analog signal may be controlled spatially under the control of enable signals provided by, or under the control of, the microprocessor 215 (e.g., a signal from a specific pixel in a specific location may be enabled). Digital data produced by the A/D converter 210 in converting an analog signal to a digital representation thereof is communicated to a data processor 225 by way of a data bus 214. The data bus 214 can be a parallel bus, a serial bus, or any other type of data bus that is convenient for moving digital data.

[0050] In one embodiment, the data processor 225 and the microprocessor 215 are the same processor, and in other embodiments they are distinct processors. Data processor 225 can be any conventional general purpose computer processing unit (CPU), programmed with suitable computer instructions, and having an associated memory, that is sufficiently powerful to manipulate the digital data in suitable time for the data analysis process. Examples of devices that can be used as data processor 225 include the Analog Devices Shark DSP, and the Texas Instruments 5510 DSP. In some embodiments, the data processor 225 and the A/D converter 210 are integrated on a single integrated circuit.

[0051] In another embodiment, optional additional components useful for testing or calibrating the circuit 200 may be present. These optional components include a source of illumination, such as a lamp 240, that provides illumination 242 that can be detected by the sensor 220. The lamp 240 optionally may be controlled by a control circuit 250 under the control of the microprocessor 215. The lamp 240 can be operated at various intensities under the control of control 250 so as to simulate different illumination levels that may be encountered in actual field operation of circuit 200. The testing and/or calibrating operation may be used in the course of manufacturing products that embody the invention, or incorporate circuits constructed or operating according to principles of the invention.

[0052] FIG. 3 is a flow diagram 300 that schematically illustrates an analog-to-digital conversion process of the invention. The process starts at oval 305, labeled “START.” The circuit 200 of the invention observes a quantity of analog data, as indicated by the box 310 labeled “OBSERVE DATA.” The data need not be, but can be, enough data to complete a reading of a decodable symbol. If the data collected at step 310 is insufficient to provide data for decoding a decodable symbol, the full set of data can be obtained either before or after the A/D converter 210 is configured to digitize the data. Alternatively, in other embodiments, a separate measurement of the ambient lighting conditions is made, for example, using the image sensor 220, or using another illumination detector such as a light meter, for example a light meter such as any of the type of light meters used in photography, with or without averaging or weighting of the illumination signal.

[0053] There are various additional ways to measure the dynamic range of the analog excitation and/or the analog signal. In one embodiment, the data processor 225 analyzes converted image data to determine a range for the operation of the A/D converter 210, and signals the microprocessor 215 to adjust the range of operation of the A/D converter 210. In another embodiment, the sensor 220 comprises a microprocessor or other data manipulation capability, and the sensor 220 signals a range to the microprocessor 215. A color CMOS M x N pixel sensor having integrated A/D converters therein is disclosed for example in Dutch Patent No. 1013911. The signal from the sensor 220 can comprise a maximum and a minimum, or the signal can be an average value observed by the sensor 220. In a further embodiment, the sensor 220 can be scanned, for example under the control of microprocessor 215, and the maximum and minimum values determined. Those of ordinary skill will envision other methods of obtaining information indicative of an operating range for the A/D converter.

[0054] In one embodiment, the circuit examines the data to determine a maximum intensity (hereinafter “max”) and a minimum intensity (hereinafter “min”) contained in the data observed by the sensor 220, as denoted by the process step represented by box 315, labeled “DETERMINE MAX, MIN.” In an alternative embodiment, a value of ambient lighting conditions observed for example by a light meter, including the “min” and the “max,” can be correlated to a range of response for the A/D converter, for example using an equation, or using a lookup table in a memory, to determine the appropriate V_ref1 and V_ref2 to be applied to the respective reference voltage input terminals of A/D converter 210.

[0055] At step 320, corresponding to the box labeled “Set V_ref1, V_ref2,” microprocessor 215 sets the values of V_ref1 and V_ref2 to prepare A/D converter 210 to perform conversions of the analog data from the sensor 220 at an appropriate range and sensitivity. As has already been described, the microprocessor 215 can prepare the A/D converter 210 for a suitable range of operation by holding one of V_ref1 and V_ref2 constant, and changing the value of the other of V_ref1 and V_ref2.

[0056] At step 325, the A/D converter 210 digitizes the analog data provided by sensor 220 according to clocking signals from microprocessor 215. When sufficient analog data has been converted, the process cycle is complete, as indicated by the oval 330 labeled “end.”

[0057] FIG. 4A is a flow diagram 400 that schematically illustrates the steps in performing the configuration of the A/D converter in overview. In FIG. 4A, the process begins at the oval 405 labeled “START.” The process includes the step of determining a value of a variable A=max−min, as indicated at box 410. As those of ordinary skill in the analytical arts will recognize, A is a measure of the dynamic range of the signal corresponding to the excitation from the encoded symbol. The process uses the value of A to adjust the behavior of the A/D converter, as indicated in the box 415. The process uses the adjusted A/D converter to analyze an observed signal, as indicated in the box 420.

[0058] FIG. 4B is a flow diagram 402 that illustrates one embodiment of the A/D configuration logic flow. In FIG. 4B, a somewhat more detailed flow logic diagram is disclosed and explained. Once again, the process starts at the oval 405 labeled “START.” The system observes at least one set of observations of a symbol to be decoded. The process includes the step of determining a value of a variable Δ=max−min, as indicated at box 410. The value of Δ is then compared to two threshold values, threshold1 and threshold2, as indicated at diamond 425. In one embodiment, threshold1 is a value below which the dynamic range of the signal (e.g., the value of a) is too small; that is, the value of the least significant bit of the A/D converter is too large. In one embodiment, threshold2 represents a value above which the dynamic range of the signal is too large, and the A/D converter will clip the signal at one or both extremities. If the value of A satisfies the relationship threshold1<Δ<threshold2, then the A/D converter is well suited to measure a signal having a dynamic range of A without clipping and without loss of sensitivity. Therefore, when the relationship threshold1<Δ<threshold2 is valid, an image that has been or that is about to be observed is suitable for analysis, as indicated at box 430, and the process is complete at oval 435 labeled “END.” The process can of course be repeated as often as is desired.

[0059] In the event that the relationship threshold1<Δ<threshold2 is not valid, one of two conditions must hold. Either Δ<threshold1, in which case the dynamic range of the signal is too small, or threshold2<Δ, in which case the dynamic range of the signal is too large. One adjusts the A/D converter to compensate for the “error condition.” In the first instance, the A/D converter is made more sensitive (i.e., the values of V_ref1 and V_ref2 are brought closer together, by adjusting at least one of V_ref1 and V_ref2). In the second instance, the A/D converter is made to span a larger dynamic range (or is equivalently made less sensitive) by adjusting at least one of V_ref1 and V_ref2 to increase the separation of V_ref1 and V_ref2. These adjustments are indicated by the box 440 labeled “Identify Vrefs based on min and max.” The values Of V_ref1 and V_ref2 can be determined by calculation, by resorting to look-up tables based on prior calibration runs, or other methods of determining what values of V_ref1 and V_ref2 will cause the relationship threshold1<Δ<threshold2 to become valid. As indicated in box 450, suitable digital values based on min and max are applied to digital-to-analog converters (D/As) (i.e., a value for D/A(min) and a value for D/A(max)) so that the desired V_ref1 and V_ref2 are applied to the A/D converter. In some embodiments, as indicated above, the D/A converters are part of the microprocessor 215. In other embodiments, the D/A converters can be provided as components of the A/D converter (e.g., as discrete V_ref1 and V_ref2 input sections thereof), so that the microprocessor 215 would provide digital signals as inputs to the V_ref1 and V_ref2 ports of the A/D, rather than voltages per se. In one embodiment, the D/As receive digital signals by way of serial communication interfaces. The process then proceeds via the step of box 430 to completion as before.

[0060] A series of tabular data showing results of actual tests are presented below. The data represent an improvement in performance, such as an increase in effective working distance, for a series of conditions. In each section of the table, the leftmost column represents a feature resolution of an encoded symbol, such as a bar code resolution in mils. The first column of data represents a nominal working distance at which the reader operates at a first value of Δ, such as Δ=1.8 Volts. In the next column of data, a relative improvement in working distance compared to the nominal value is presented, for a second value of Δ, such as a nominal 1.0 Volts. The data show clear improvements for low levels of illumination. It will be understood that improvements are also possible for higher levels of illumination. However, at high levels of illumination the performance of the unoptimized reader is sufficiently good that the improvements to be gained will in general be small. The reason for the smaller improvements at high illumination levels is that there already is enough light to obtain a suitable image and to decode the image with high accuracy and precision even when the operation of the A/D converter is somewhat less than optimal. Nevertheless, there may be situations in which optimal performance even at the highest illumination levels, such as noonday sun on a clear summer day, or approximately equal to 70,000 Lux, will be required or desired. The tabular data is now presented. 1

[0061] For the example of a non-coated window, zero Lux ambient illumination, and line resolutions in the range of 13 to 55 mils, the improvement that is realized ranges from 15.7% to 28.5%. 2

[0062] For the example of a non-coated window, 25 Lux ambient illumination, and line resolutions in the range of 13 to 55 mils, the improvement that is realized ranges from 3.8% to 33.7%. 3

[0063] For the example of disruptive optics, zero Lux ambient illumination, and line resolutions in the range of 13 to 55 mils, the improvement that is realized ranges from 15.8% to 34.6%. 4

[0064] For the example of disruptive optics, 25 Lux ambient illumination, and line resolutions in the range of 13 to 55 mils, the improvement that is realized ranges from 4.2% to 26.7%. Therefore, as shown in the above experimental observations, in general one can except that the replication of the systems and methods of the invention provide an improved signal processing circumstance that offers the advantages of both improved accuracy and improved precision. These advantages are also observed in prototype devices that embody principles of the invention. It will be apparent to those of ordinary skill that the principles of the invention apply equally to other embodiments employing A/D converters for use in many other applications, in addition to the reading and decoding of encoded symbols such as bar codes. For example, the principles of the invention can be applied to any analog signal that is subject to a varying dynamic range, such as optical or other electromagnetic signals, audible signals, thermal signals, pressure and/or vacuum signals, and others, such as signals derived from the chemical interaction of substances with sensors.

[0065] Those of ordinary skill will recognize that many functions of electrical and electronic apparatus can be implemented in hardware (for example, hard-wired logic), in software (for example, logic encoded in a program operating on a general purpose processor), and in firmware (for example, logic encoded in a non-volatile memory that is invoked for operation on a processor as required). The present invention contemplates the substitution of one implementation of hardware, firmware and software for another implementation of the equivalent functionality using a different one of hardware, firmware and software. To the extent that an implementation can be represented mathematically by a transfer function, that is, a specified response is generated at an output terminal for a specific excitation applied to an input terminal of a “black box” exhibiting the transfer function, any implementation of the transfer function, including any combination of hardware, firmware and software implementations of portions or segments of the transfer function, is contemplated herein.

[0066] While the present invention has been explained with reference to the structure disclosed herein, it is not confined to the details set forth and this invention is intended to cover any modifications and changes as may come within the scope of the following claims.