GB2200778A

GB2200778A - Object identification

Info

Publication number: GB2200778A
Application number: GB08702473A
Authority: GB
Inventors: Philip Edward Stephens
Original assignee: General Electric Company PLC
Current assignee: General Electric Company PLC
Priority date: 1987-02-04
Filing date: 1987-02-04
Publication date: 1988-08-10
Anticipated expiration: 2007-02-04
Also published as: GB2200778B; GB8702473D0

Abstract

A system classifies objects such as coins by driving them into mechanical resonance and then detecting the frequencies and relative amplitudes of some of the resonant frequencies of the object. This detection is carried out by using a microphone 5 to listen to the sound produced by the object. <IMAGE>

Description

OBJECT IDENTIFICATION This invention is concerned with apparatus for classifying objects.

It is often useful to classify an object as being of one of a number of possible types of object. One example of a situation where this is necessary is in a coin-freed vending machine. In such machines it is essential to identify whether an object supplied to the machine is a valid coin or not. In many such machines it is desirable to be able to distinguish several different types of valid coins from one another and from other objects. Another example is found in the manufacture of metal components where it is necessary to examine the components and classify them as acceptable or unacceptable, depending on whether or not their shape and size is within a small range of values.

At present such apparatus identifies coins or classifies components-by measuring their dimensions and physical qualities such as weight, and other features such as the presence of a knurled rim. Each valid coin type or classification of component will have a set of values for these qualities so the identity of each coin or classification of each object can be deduced from its values for these qualities.

Such systems have a number of drawbacks however.

Being mechanically complex, they are relatively expensive and subject to failure due to wear,- externally produced vibration and accumulated dirt from the coins or objects.

It was the object of the present invention to produce an object identification system that at least partially overcomes these problems.

This invention provides apparatus for classifying objects including means for sensing resonant frequencies of the objects.

Two embodiments of the invention are now described with reference to the accompanying drawings in which: Figure 1 shows a coin identification mechanism constructed in accordance with the invention in schematic form; and Figure 2 shows a quality-control mechanism constructed in accordance with the invention in schematic form; identical components having the same reference numerals throughout.

Referring to Figure 1 in a vending machine a coin 1 is inserted through a slot 2 by a customer. The term coin is used to mean any coin or object similar enough to a coin to be possibly mistaken for one. The coin then rolls down a chute 3. The coin 1 drops from the end of the chute 3 onto a metal block 4. The coin 1 will always land at the same point on the block 4, this position being determined by the profile and position of the chute 3. A coin 1A is shown in this position.

A microphone 5 is positioned adjacent to the coin landing point, and a piezo-electric transducer 6 is attached to the bottom of the block 4.

When a coin 1 lands on the metal block 4 the coin 1 will vibrate at its resonant frequencies and a shock wave will be generated in the block 4. The shock wave in the block 4 will cause the piezo-electric transducer 6 to generate a signal while the vibration of the coin 1 causes the microphone 5 to produce signals having relative amplitudes and frequencies equivalent to those of the sounds produced by the vibration of the coin 1.

Since all coins 1 will follow the same path from the end of the chute 3 to the block 4 the size of the shock wave generated in the block 4 will be related to the mass of the coin 1. Thus the amplitude of the electrical signal produced by the piezo-electric transducer 6 will also be related to the mass of the coin 1. Each type of coin 1 will have a set of resonant frequencies dependent on the diameter, thickness and metallurgical make up of the coin 1, and also on the shape of non-circular coins.

The value of coin 1 can be deduced from its mass and resonant frequencies.

The signals produced by the microphone 5 are supplied to a number of bandpass filters 7A, 7B, etc. to 7N.

Each bandpass filter has a different pass band including at least one resonant frequency of one type of valid coin.

The filtered signals from bandpass filters 7A to 7N and the signals produced by the piezo-electric transducer 6 are fed to the processor 8.

It will usually be the case for a set of national currency that some of these resonant frequencies will be shared by more than one denomination of coin, so in order to unambiguously identify a coin it is necessary to look at a number of resonant frequencies and measure their relative amplitudes.

The processor 8 u,ses these signals to determine whether the coin 1 is a valid acceptable coin or not and if it is what value coin it is by measuring the amplitude of the signal produced by the transducer 6 and by comparing the relative amplitudes of the signal passed by the filters 7A to 7N and then comparing these with a look up table stored in a memory 18. In response to instructions found in this look up table, the processor 8 produces signals on a line 9 to control the destination of the coin 1, which may be a number of storage areas or a reject slot (not shown) depending on the coins validity and value and produces signals on a line 10 which inform a processor controlling the vending machine of the value of the coin so that it can control vending accordingly.

In order to reduce the power consumption of the system the processor 8 can be normally switched off except for a rising edge detector looking for a rising edge in the signals from the piezo-electric transducer 6 and arranged to switch on the rest of the system when such a rising edge is detected.

In Figure 2 a manufacturing quality control system is shown which checks the accuracy of manufacture of castings 11.

The castings 11 are carried along a conveyor 12 which drops them onto a metal block 4. The conveyor 12 and the block 4 are arranged so that all of the castings 11 will strike the metal block 4 in the same spot with the same orientation.

A microphone 5 is placed adjacent to the spot where the castings 11 will land and a piezo-electric transducer 6 is attached to the bottom of the block 4. When a casting 11 strikes the block 4 a shock wave is produced in the block 4, and the casting 11 is caused to resonate, the vibrations produced by the resonating casting 11 are picked up by the microphone and the shock wave in the block 4 is picked up by the piezo-electric transducer 6.

The signals produced by the microphone -5 and the transducer 6 are supplied to a processor 13. On receiving a signal from the transducer 6 the processor 13 prepares to receive a signal from the microphone 5, the processor 13 digitally records a portion of the signals produced by the microphone 5.

The signals recorded by the processor 13 are then compared by a comparator 14 with signals recorded in a memory 15. The signals stored in the memory 15 are those produced when a perfect casting is dropped onto the block 4. These signals are placed in the memory 15 during the setting up of the system by the processor 13 feeding the signals it records to the memory 15,instead of the comparator 14, when an ideal or perfect casting that has been checked by direct physical measurement is dropped onto the block 4.

If the correlation between the signals recorded - by the processor 13 and those stored in the memory 15 is above a preset level the comparator 14 signals along a line 16 and the casting 11 continues the manufacturing process, if the correlation is not above this level the comparator 14 signals along a line 17 and the casting 11 is diverted to a reject bin. Machinery for carrying out such routing of components is well known and will not be described here.

The analysis of the resonant frequencies in the systems described could alternatively be carried out by a fast fourier transform technique.

The driving of objects to resonance could alternatively be carried out in many ways, other than described, for example phot-accoustically by means of a modulated laser beam.

Although this invention is described with reference to coins and castings it can be applied to any objects capable of producing a resonant frequency.

Claims

1. Apparatus for classifying an object including means for sensing a resonant frequency of the object.

2. Apparatus as claimed in claim 1 and wherein said means are responsive to a predetermined value of a resonant frequency

3. Apparatus as claimed in claims 1 or 2 and including means to drive the object into resonance.

4. Apparatus as claim in claim 1, 2 or 3 and including means operative to cause the object to strike a relatively massive body in a predetermined manner, and means adapted to sense the response of said body to said object.

5. Apparatus as claimed in claim 4 and wherein the means which drives the object into resonance includes the relatively massive body.

6. Apparatus as claimed in claim 5 and wherein the relatively massive body is a metal block.

7. Apparatus as claimed in any of the preceeding claims and wherein means are provided for identifying a plurality of resonant frequencies associated with a predetermined object.

8. Apparatus as claimed in claim 7 and wherein a plurality of different predetermined objects are classified by identifying a plurality of resonant frequencies which uniquely characterise each class of object.

9. Apparatus as claimed in claim 7 or 8 and wherein the relative amplitudes of different resonant frequencies of said plurality are additionally utilised to classify an object.

10. Apparatus as claimed in claim 7, 8 or 9 and wherein a set of bandpass filters are provided to facilitate the identification of predetermind resonant frequencies.

11. Apparatus as claimed in any of the preceding claims and wherin the means for sensing a resonant frequency of the object includes a microphone.

12. Apparatus as claimed in any of claims 4 to 11 and means responsive to the magnitude of a shock wave induced by the object in said body to further facilitate the classification of the object.

13. Apparatus as claimed in claim 12 and wherein a piezoelectric transducer in contact with said body is used to sense the shock wave.

14. Apparatus for classifying objects substatially as shown in and substantially as described with reference to figure 2 of the accompanying drawings.

14. Apparatus as claimed in claim 13 and wherein means for classifying the object is activated when said transducer generates a signal in response to the presence of an object.

15. Apparatus as claimed in any preceding claim where said objects are coins.

16. Apparatus for classifying coins substantially as shown in and substantially as described with reference to figure 1 of the accompanying drawings.

17. Apparatus for classifying objects substantially as shown in and substantially as described with reference to figure 2 of the accompanying drawings.

Amendments to the claims have been filed as follows CLAIMS 1. Apparatus for classifying an object as being in one or none of a plurality of different classes of objects, each class of object being uniquely distinguished by a resonant frequency; including means for driving the object into resonance, means for sensing a resonance frequency of the object including means for sensing a resonant frequency of the object and means for comparing the resonant frequency of the objects with the resonant frequencies of the classes of objects.

2. Apparatus as claimed in claim 1 in which each class of objects is uniquely distinguished by a set of one or more resonance frequencies and the apparatus includes means for sensing a plurality of resonance frequency or frequencies of the object with the sets of resonant frequencies of the classes of objects.

3. Apparatus as claimed in claims 2 in which the relative amplitudes of the different resonance frequencies in a set are used to distinguish each class of objects; including means sensitive to the relative amplitudes of the resonant frequencies of the object and means for comparing the relative amplitudes of the resonant frequencies of the object with the relative amplitudes of the resonance frequencies in a set.

4. Apparatus as claim in claim 1, 2 or 3 and including means operative to drive the object into resonance by striking a relatively massive body in a predetermined manner, and means adapted to sense the response of said body to said object.

5. Apparatus as claimed in claim 4 and wherein the relatively massive body is a metal block.

6. Apparatus as claimed in claims 2 and 5 and wherein a set of bandpass filters are provided to facilitate the identification of predetermined resonant frequencies.

7. Apparatus as claimed in any of the preceeding claims and wherein means for sensing a resonant frquency of the object includes a microphone.

8. Apparatus as claimed in claims 4 to 7 and including means responsive to the magnitude of a shock wave induced by the object in said relatively massive body to further facilitate the classification of the object.

9. Apparatus as claimed in claim 8 and wherein a piezoelectric transducer in contact with said body is used to sense the shock wave.

10. Apparatus as claimed in claim 9 and wherein means for classifying the object is activated when said transducer generates a signal in response to the presence of an object.

11. Apparatus as claimed in any preceding claim where said objects are coins.

12. Apparatus for classifying an object as being in one or none of a plurality of different classes of objects, each class of object being distinguished by frequency charateristics of its resonation; including means for causing the object to resonate, means for sensing the frequency characteristics of the object and means for comparing the sensed characteristics with stored characteristics of the said different classes of objects.

13. Apparatus for classifying coins substantially as shown in and substantially as described with reference to figure 1 of the accompanying drawings.