IE50163B1 - Improved echo recognition system - Google Patents

Abstract

An echo type rangefinding a system, preferably using ultrasonic energy bursts, includes a transducer for transmitting and receiving the energy burst and signal processing means for identifying a true echo of the energy burst (54) from a selected target. The signal processing means discriminates against spurious signals and electronic noise, some of which may be generated by the rangefinding system itself, by rejecting signals that do not persist for at least a predetermined period. Preferably, the system recognizes the leading edge of a true echo signal by sampling the integrated signal (156) at sampling intervals (158) which are small compared with the said predetermined period (for example 0.6 milliseconds). If the samples integrated signal voltage (Vc) steadily increases at at least a given minimum rate for at least the predetermined period then the signal processing means provides an "object detect" signal. The accuracy of the rangefinding system is improved firstly because of the improved signal discrimination, and secondly because the object detect signal is always produced at the same time after the beginning of the true echo signal.

Description

The present invention relates to an echo-type range finding system which discriminates from spurious signals and/or electronic noise and has an improved distancedetermining accuracy.

Xn a known camera employing an ultrasonic range finding system, a reflection or echo of a portion of a burst of ultrasonic energy, previously transmitted by the camera, is sensed by the camera's range finding system for the purpose of regulating the camera's automatic focusing system in accordance with a signal representative of the total flight time of said burst of ultrasonic energy. Such rangefinding systems are generally susceptible to spurious signals and/or electronic noise sensed or generated by the rangefinding system. If a spurious signal is sensed by a camera's rangefinding system after the transmission of a burst of object detecting energy, but before a true or actual object echo can be received by the system, a misfocusing of the camera's lens by said automatic focusing system would result.

In order to reduce the sensitivity of the rangefinding system to spurious signals and/or electronic noise, all received signals above a threshold level are integrated by an integrating capacitor. An object detect signal is not generated until the voltage of this capacitor exceeds a predetermined trigger level. However, the duration and shape of a received object detection signal is dependent upon several variables that include object distance, object shape, path length differences of portions of a reflected signal, etc., and therefore, the use of such an integrator to reduce sensitivity to spurious signals can introduce distance errors into the range finding system.

Spurious signals can emanate from a number of different sources. In, for example, the rangefinding system described, ultrasonic energy is both transmitted and received by a combination transmitting and receiving, capacitance-type, electrostatic transducer. This type of transducer includes a vibratile diaphragm that vibrates and transmits a burst of ultrasonic energy in response to a series of high frequency input signals, and also vibrates when it receives an echo of a previously transmitted burst of ultrasonic energy, causing the transducer to generate a relatively low level receive or object detect signal at its output.

This type of transducer has common input and output signal terminals and, therefore, the circuitry that respond to a receive signal must be blanked or rendered insensitive to signals appearing at the transducer input/output terminals during the transmit mode. This receive circuitry blanking means is removed sometime after the transmit signal Is terminated. In some transducers, the vibrating diaphragm will break into oscillations or beat after diaphragm vibrations have fully decayed and after the receive circuitry has been unblanked or made sensitive to signals appearing at the transducer input/output terminals. This diaphragm beating is one type of spurious object detection signal that can, for example, cause the lens misfocusing problem mentioned above.

The object detection system described above using sonar range-finding for an automatic focusing camera also includes a variable gain receiver amplifier whose gain is changed, in a series of discrete steps, as a function of the time of flight of a burst of ultrasonic energy. Amplifier gain is increased in this manner to compensate for the difference in magnitude between the echo signals received from close objects and the substantially less intense echo signals received from remote objects. When the amplifier gain is changed in a series of steps, as mentioned above, electronic noise, of relatively short duration, appears from time to time at these gain step change points, thereby producing a source of electronic noise that can cause false triggering and therefore misfocusing of the camera's automatic focusing system.

The amplitude of all of the spurious signals and electronic noise described above are time-dependent, and exist for a relatively short period of time. For 20 example, the spurious signals generated by the abovedescribed transducer beating phenomenon and the electronic noise generated by the receiver amplifier at its gain step change points are in the order of 200 microseconds .

The present invention consists in a rangefinding system having means for transmitting a burst of energy of predetermined duration towards a target and for receiving, and distinguishing from spurious signals, an echo of the burst, comprising: sampling and summing means providing signals representing the sum of the magnitudes of consecutive portions of the received signals at intervals of short duration relative to the duration of the echo; and discriminating means for accepting the received signals as an echo only if the sum of the magnitudes of the said portions increases at at least a predetermined rate over a predetermined period.

Preferably, the discriminating means for accepting the echo includes means for resetting the said sum to an initial value each time the sum fails to increase or the rate of increase, during any given interval, is less than the said predetermined rate.

The rangefinding system of the present invention achieves good discrimination against spurious signals and/or electronic noise. Object distance errors are readily removable by electronic or mechanical means because the error or time delay introduced into the rangefinding system by the above-mentioned sampling scheme is the same for all detected objects.

Xn order that the invention may be better understood, a known rangefinding system and a rangefinding system embodying the present invention will now be described with reference to the accompanying drawings, wherein:Figure 1 is a block diagram that primarily shows the spurious signal rejection portion of an ultrasonic range finding system in accordance with the prior art; Figure 2A shows a typical transmit/receive signal that initially appeared at the input/output terminals of an electrostatic transducer in an ultrasonic range finder, after said signal has been amplified; Figure 2B is a graph of voltage as a function of time on the spurious signal rejecting integrating capacitor in the prior art ultrasonic range finder of Figure 1; Figure 2C is a graph of voltage as a function of time at the output of the switching means in Figure 1 that is responsive to the integrating capacitor voltage depicted in Figure 2B; Figure 3 is an enlarged detail of a transmit/ receive signal and of the voltage on. an integrating capacitor similar to that depicted in Figures 2A and 2B, respectively; Figure 4A is a representation of an actual object detection signal returning from a relatively close object initially appearing at the input/output terminals of an electrostatic transducer in an ultrasonic rangefinder, after said signal has been amplified; Figure 4B is a representation of an actual object detection signal returning from a relatively remote object initially appearing at the input/output terminals of an electrostatic transducer in an ultrasonic rangefinder, after said signal has been amplified; Figure 5 is an enlarged detail of two object detection signals having the same frequency, but having substantially different magnitudes, that are superimposed on one another; Figure 6 is a block diagram that primarily shows the spurious signal rejection portion of an ultrasonic xangefinding system in accordance with the present invention; Figure 7 is a logic flow diagram of the echo recognition system of the present invention; 30163 Figure 8 is primarily a representation of the returning or echo portion of an object detecting signal and a graph of the voltage corresponding to said signal on the small interval voltage summing capacitor of the present invention, as a function of time.

As ultrasonic rangefinders are known, only the spurious signal rejecting means in the ultrasonic rangefinder, a block diagram of which is shown in Figure 1, will be discussed in detail.

Referring now to Figure 1 and to the prior art spurious signal rejection portion of an ultrasonic rangefinder depicted therein, system time base generator or clock 12 is connected to a battery (not shown) attached to terminal 14 through two-position manuallyactuable switch 16. Wk The receive portion of the rangefinder is connected to transducer 20 through transmit and blanking generator 18, a full wave rectifier (not shown) and path 24. Inasmuch as the input and output terminals of transducer 20 are common to one another, it is essential that all transmit signals be prevented from entering path 24 so that the rangefinder receiver will not confuse a transmit signal with an echo of said transmit signal. This is accomplished by blanking circuitry in transmit and blanking generator 18 that prevents signals from entering path 24 while a transmit signal is present at the input/output terminals of transducer 20 and, because of transmit signal decay time, for a short time thereafter.

When path 24 is not blanked, an electrical signal generated by transducer 20 upon receipt of an echo of a previously transmitted burst of ultrasonic energy, is routed to path 24 through a full wave rectifier (not shown) and transmit and blanking generator 18. When the receive signal on path 24 equals or exceeds a predetermined magnitude, as determined by level detector 26, said detector 26 turns on gate 28, thereby connecting integrator capacitor 30 to constant current source 32. When the magnitude of the voltage on integrating capacitor 30 equals or exceeds a predetermined magnitude, as determined by level detector 34, switch means 36, in the form of, for example, a schmitt trigger, conducts and generates object detect signal 38 at its output. Object detect signal 38 is subsequently combined with other rangefinder signals to determine object distance, an operation that is not essential to the operation of the above-described spurious signal rejecting means. A better understanding of the operation of the prior art spurious signal rejecting means of Figure 1 may be acquired by additionally referring to the typical transmit/receive object detecting signal shown in Figure 2A and how the voltage on the integrating capacitor incorporated in said signal rejecting means changes in response to said signal, as shown in Figure 2B, to generate the object step voltage shown in Figure 2C. Figure 2A is an oscilloscope trace of an actual transmit/receive object detection signal initially appearing at the input/output terminals of an electrostatic transducer 20 (Figure 1), after said signal has been amplified.

Multiple frequency transmit signal 40 having several frequencies in the neighbourhood of 50-60 KHz is applied to the input/output terminals of transducer 20 for 1.1 ms. Vibrations of the diaphragm in transducer 20 fully decay in about an additional 0.3 ms. Previously explained transducer beating occurred at 4 2 and 44 after the vibrations in the diaphragm of transducer 20 have fully decayed. The initial beating at 42 is less than the threshold level established by level detector 26 and therefore the voltage on the integrating capacitor in integrator 30 remains at its initial zero value. Even if the magnitude of the voltage resulting from beating 42 exceeded threshold level 46, very little of it would have caused the charging of integrator 30 from constant current source 32 because the input to the rangefinder receiver was “blanked11 for a total of 1.6 ms. Blanking renders the range finder receiver insensitive to all signals appearing at the input/output terminals of transducer 20. A second beating occurs at 44 and the magnitude of the voltage resulting from saidbeating“ exceeds threshold level 46 established by level detector 26, causing integrator 30 to be charged from constant current source 32 for the period of time that the voltage resulting from beating 44 exceeds threshold level 46.

In this instance, the voltage on the integrating capacitor and integrator 30 increases, but peaks at 48, ίο slightly below object detect or trigger level 50, a level that is established by level detector 34, and then linearly decreases to zero without generating object detect signal 38. Additional spurious signals caused by reflections from off-axis objects within the side lobes of transducer 20 that are closer to the transducer than the main target, appear at 52 but their magnitudes are less than threshold level 46. Finally, actual receive signal 54 exceeds threshold level 46 established by level detector 26 for a sufficient period of time to cause the voltage on the integrating capacitor in integrator 30 to reach object detect or trigger level 50 at 56, causing level detector 34 to actuate switching measn 36 to its conducting state and generate object detect signal 38 at its output.

The integrating capacitor in integrator 30 integrates whenever the amplified voltage resulting from either an actual or apparent object detection signal exceeds threshold level 46, as previously explained. Whenever the magnitude of the apparent or actual received signal voltage drops below threshold level 46, the voltage on said integrating capacitor linearly decays toward zero. The voltage on the integrating capacitor of integrator 30 and a time 25 expanded portion of a receive signal that results in such a capacitor voltage are shown in Figure 3. During times 58 and 60, the amplified received signal voltage equals or exceeds threshold level 62 and causes the integrating capacitor to linearly increase as it is charged frorr. constant current source 32. At all other tiir.es, the integrating capacitor voltage either decreases to or remains at the zero voltage level. Because of the relatively slow decay time associated with integrating capacitor 30 (Figure 1) after a spurious signal drops below threshold level 62, said threshold level 62 tends to be relatively high in order to minimize integrating capacitor charging by a spurious signal that might combine with a subsequent spurious signal to cause false object detection signal triggering.

In addition, typical amplified receive signals from a target objects may look like those depicted in Figures 4A or 4B. Receive signal 64 in figure 4A starts off and remains at a relatively high average value while receive signal 66 in Figure 4B starts off at a relatively low average value and then increases to a relatively high average value. Signal 64 is characteristic of an echo from a close object whereas signal 66 is characteristic of an echo from a relatively remote object. This difference in echo or receive signal magnitude could cause integrator 30 to start integrating sooner or later, depending upon signal shape, which introduces variable, distance determining errors into the rangefinding system. These errors are of more concern for low level signals near the threshold level established by level detector 26 than they are for signals of substantially greater magnitude. The reason for these errors can be readily seen by referring to the two, time-expanded, receive signals superimposed on one another in Figure 5.

In Figure 5, a single cycle of receive signals 68 and 70 having the same frequency are artificially superimposed on one another to facilitate explaining why receive or echo signals having substantially different magnitudes introduce distance determining errors into a range system having the spurious signal rejecting means of Figure 1. When signal 68 equals or S0163 or exceeds threshold level 72, a level that would, for example, be established by level detector 26 in Figure 1, the Integrating capacitor in integrator 30 would be charged at a constant rate for a period of time corresponding to time 74. However, signal 70 whose magnitude is substantially greater than that of signal 68, could cause the integrating capacitor in integrator 30 to charge for a longer period of time or the period of time corresponding to time 76. This difference in charging time causes an integrating capacitor voltage difference that renders the accuracy of the measured object distance dependent upon signal magnitude.

Turning now to the present invention and specifically to Figure 6 of the drawings, a block diagram showing the spurious signal and electronic noise rejecting portion of an ultrasonic range finding system in accordance with the teachings of the present invention, is depicted. In Figure 6, system time base generator or clock 78 is connected to a battery 20 (not shown) attached to terminal 80 through two-position manually actuable switch 82. When switch 82 is actuated to its closed position, a high frequency oscillator in system time base generator 78 is energized and its divided-down output provides the time base or reference for all time-related rangefinder functions.

The closure of switch 82 also connects other rangefinder system components to terminal 80. After system time base generator 78 has been energized, its output is routed to transmit and blanking generator 84 causing said generator 84 to apply the appropriate signal to the input/output terminals of electrostatic transducer 86 through path 88, causing said transducer 86 to transmit a burst of ultrasonic energy toward an object to be detected. The receiver portion of the rangefinder is connected to transducer 86 through path 90, a full wave rectifier (not shown) and transmit and blanking generator 84. As previously explained with respect to the prior art spurious signal rejecting means of Figure 1, it is essential that all transmit signals appearing at the input/output terminals of transducer 86 be prevented from entering path 90 to avoid confusing the range finder receiver. This is accomplished by blanking circuitry in transmit and blanking generator 84 that functions in the same manner as the blanking circuitry in transmit and blanking generator 18 in prior art Figure 1.

When path 90 is unblanked an electrical signal is generated by transducer 86 upon receipt of an echo of a previously transmitted burst of ultrasonic energy and this signal, after being rectified, is routed to path 90 through transmit and blanking generator 84.

The rectified signal is routed to a capacitor in sample and hold means 94 where it is continuously sampled during time intervals that are small relative to overall receive signal duration (e.g., every 25 microseconds during a 1.1 millisecond duration receive signal).

The just-mentioned capacitor in sample and hold means 94 is an integrating capacitor that integrates or sums sampled received signal voltages. Continuous signal presence is established if the magnitude of the sum of the sampled voltages always increases between successive samples and if the rate of increase is equal to or greater than a predetermined rate for the entire duration of a major portion of the receive signal.

If the magnitude of the sum of the sampled voltages on the integrating capacitor in the sample and hold means S4 is always increasing, as determined byAV magnitude determining means 96, and the rate of change of said sampled and summed voltage is equal to or greater than a predetermined rate as determined by Δ? rate determining means 98, AND gate 100 will conduct and satisfy one of the two necessary inputs to AND gate 102 When transmit and blanking generator 84 unblanked path 90, as previously discussed, an interval timer initiating signal was set through path 104 to initiate timing by interval timer 106 which for this particular 1θ rangefinder is 0.6 ms. If AND gate 100 continues to conduct for 0.6 ms, AND gate 102 will conduct when interval timer 106 produces a voltage at its output 0.6 ms from the time that a timer initiating signal was sent through path 104. When AND gate 102 conducts, switching means 108 will conduct and generate object detect signal 110 at its output. On the other hand, if either the sampled voltage magnitude fails to increase between successive samples as determined by 96 or if the rate of change of the sample voltage was less than a predetermined minimum rate, NAND gate 112 would conduct and send an initializing signal to sample and hold means 94 and interval timer 106 through path 114. This initializing signal would both reduce the voltage on the integrating capacitor in sample and hold means 94 to zero and reset interval timer 106 to zero, and then await the next interval timer initiating unblanking signal.

A more detailed explanation of how the spurious signal and electronic noise rejecting means of the present invention functions will now be decriied with respect to the logic flow diagram in Figure 7, the spurious signal and electronic rejecting system block diagram of Figure 6, and the return or echo portion of an object detection signal and the graph of integrating capacitor voltage of Figure 8.

Turning now to Figure 6, 7 and 8, in Figure 8, upper trace 54 is primarily that of a receive signal of 1.1 ms duration. The lower portion of Figure 8 is a graph of the integrating capacitor voltage in sample and hold means 94 as a function of time and receive signal strength. With reference to Figure 7, the spurious signal rejecting means of the present invention is initialized by setting interval timer 106 (Figure 6) to zero which corresponds to step (118) in the flow diagram of said Figure 7, along with a storage location in, for example, a digital computer that stores information corresponding to the magnitude of the summed voltage (V^) on the integrating capacitor in sample and hold means 94 during the previous sampling interval, said voltage corresponding to step (120). Stored (which was set to zero at [120]) is subtracted from integrating capacitor voltage V which yields a/Sv = V -V (134). The storage location of V is now set to the new V (136) which, P P as stated sbove, is the magnitude of the summed voltage on the integrating capacitor in sample-and-hold means 94. Δν is tested against a small value (close to zero) (TOL) (138) as to its magnitude. If Av does not exceed TOL (140) it means that νθ did not increase fast enough over the last sampled interval and that the apparent receive signal has decayed into the background noise level or close to it. True receive signals beat, but neverdecay into the background noise level during the 0.6 ms or more of return signal passage. Therefore, if Δν is less than TOL (140), we do not have a true signal and the voltage on the integrating capacitor Vc fi is set to zero (128) . We then wait for the next time interval (130) and initialize (118) and (120) and start over again. If, on the other hand,Av is larger than TOL (142), the time duration of the sampled interval is added to the interval timer (144) (and 106 in Figure 6) and its elapsed time from unblanking” is sampled to see if it exceeds 0.6 ms (146). If it does not exceed 0.6 ms (148), we wait for the next time interval (150). If it does exceed 0.6 ms (152), we generate true object detect signal 154.

In Figure 8, the receive signal portion 54 of the object detection signal of Figure IA, along with a graph of the integrating capacitor voltage 156 of the integrating capacitor in sample and hold means 94 (Figure 6) is illustrated. The charging and discharging of the integrating capacitor in sample-and-hold means 94 (Figure 6) is shown at 158. When the integrating capacitor voltage increases, we must determine whether this signal is spurious or a true signal. Spurious signals do not usually persist for 0.6 ms. in the present system. Therefore, when the slope of the integrated signal goes to TOL/ AT (Δτ being equal in time to a sampling interval), or if the slope is less than a predetermined one (Λν less than TOL) (140) , the system is initialized (118) and (120) after the integrating capacitor is discharged (128) and after the sampling interval has passed (130).

Prior art rangefinding systems of the type schematically depicted in Figure 1 rely on signal shape or amplitude for target signal recognition. Unlike such prior art systems, the present system will detect a signal of energy returning from a selected target for any object, signal strength or pulse shape. The time of the occurrence of the leading edge of a signal of energy or true echo can readily be established, to within one sampling interval, by determining the elapsed time between the first sampling interval of an actual or true echo from a selected target and the generation of an actual or true object detect signal and then subtracting said elapsed time from the time that said actual or true object detect signal is generated.

Beating signal 44 in Figure 2A could well look like a true signal at the outset. However, since beat pulses do not usually exceed 140 microseconds, and electrical spikes do not tend to exceed 300 microseconds, the change in voltage between adjacent sampling intervals will be less than TOL in less than the 0.6 ms time interval of the present system. This will cause either beat pulses or electrical spikes to fail the test at (138) and take the logic path by way of 140 to set V to zero and start over aaain c (initialize). Thus, the primary feature of a range finding system incorporating an embodiment of the present invention is its immunity to beating and noise spikes (electrical and mechanical). False triggering is avoided because these types of signals cannot pass the 0.6 ms test. Rangefinding system accuracy is improved because a true echo signal is verified, in part, by its duration and therefore the range finding system is not dependent upon gain or signal shape as to when it triggers.

The typical echo signal shown in Figure 2A, of the rangefinding system of the present invention, is of 1.1 - 1.5 ms. duration. For an echo signal of this duration, it has been empirically determined that if such a signal persists for 0.6 ms, this period of time is long enough to establish that the received signal is a true or actual echo signal and not a spurious one. The time duration of the total number of sampling intervals will be equal to this empirically determined length of time. This 0.6 millisecond period of time during which a received signal must persist in order to satisfy one of the test criteria of the echo recognition system of the present invention is not necessarily the same for all rangefinding systems utilizing such a recognition system. However, establishing a time period during which an echo signal must continuously persist is essential for the proper operation of the present echo recognition system.

When a true signal arrives, Δν is greater than TOL (142) and remains so for more than 0.6 ms (144), (146) and (152). This causes the system to generate a true object detect signal (154) and a subsequent range signal that is always 0.6 ms closer than the object detect signal (154) indicates. This is a constant error of about 15 cm and can be readily subtracted out of the ranging system (either mechanically or electronically) to consistently, although indirectly, identify the leading edge of an electrical signal.

Because of its superior spurious signal rejecting capability, the present system can look at all received signals and not only those above a particular threshold level as in prior art arrangements, which further improves rangefinder distance determining accuracy.

Claims (9)

1. A rangefinding system having means for transmitting a burst of energy of predetermined duration towards a target and for receiving, and distinguishing from spurious signals, an echo of the burst, comprising; sampling and summing means providing signals representing the sum of the magnitudes of consecutive portions of the received signals at intervals of short duration relative to the duration of the echo; and discriminating means for accepting the received signals as an echo only if the sum of the magnitudes of the said portions increases at at least a predetermined rate over a predetermined period.

2. A rangefinding system in accordance with claim 1, wherein the discriminating means for accepting the echo includes means for resetting the said sum to an initial value each time the sum fails to increase or the rate of increase, during any given interval, is less than the said predetermined rate.

3. A rangefinding system in accordance with claim 1 or 2, including means for determining the time of occurrence of the leading edge of the echo by subtracting the predetermined time from the time at which the echo is identified by the discriminating means.

4. A rangefinding system in accordance with claim 1, 2 or 3, wherein the sampling and summing means includes an integrating capacitor to which the consecutive portions of the received signals are applied.

5. A rangefinding system in accordance with claim 4, wherein the discriminating means includes comparator means storing the value of the voltage on the integrating capacitor for the preceding interval and 5 providing a signal representing the difference between the voltage level on the integrating capacitor for the current interval and the stored level.

6. A rangefinding system in accordance with claim 5, wherein the discriminating means further includes 10 means responsive to the comparator means for determining whether the said voltage level for the current interval exceeds the stored voltage level by a difference greater than a predetermined value, and operative to reset the capacitor voltage to zero if 15 it does not.

7. A rangefinding system according to any one of the preceding claims, wherein the burst of energy comprises a burst of sonic energy.

8. A rangefinding system according 20 to claim 7, wherein the predetermined period is substantially 0.6 ms.

9. A rangefinding system substantially as herein described with reference to Figures 7 and 8 of the accompanying drawings.