CN105158818A - Signal evaluation method of X-ray detection device - Google Patents

Abstract

The invention provides a signal sampling and evaluation method of an X-ray machine. The method comprises: step one, before a machine leaves a factory, the performance of the machine is guaranteed to have reached a stable and normal state, a detector signal is sampled under a specific machine parameter and is stored to form a standard signal base; step two, after the machine is started to be used, a detector terminal signal is sampled regularly; and during signal sampling, the parameter of the machine is required to be consistent with the machine parameter when the standard signal base is established; and step three, a collected detector terminal signal sample is compared with a standard signal in the standard signal base and whether the signal value is in a reasonable range is evaluated. The method has the following advantages: a signal received by a detector can be examined automatically after the machine is started each time; a scientific quantitative criteria is set; a machine problem can be found timely based on the signal change; and a problem that the machine detection performance becomes poor because the machine can not be examined regularly can be solved.

Description

A kind of method of X-ray detection equipment signal evaluation

Technical field

The present invention relates to a kind of signal evaluation method, in particular a kind of method of X-ray detection equipment signal evaluation.

Background technology

The general structure of current X-ray equipment for detecting foreign matter is: the source of penetrating is positioned at above belt, and detector is positioned at below belt, material by belt transport, material through detector position, detector tracer signal.This signal is identified by special processor, thus detects the foreign matter in material.

But along with the Long-Time Service of X-ray machine, penetrating source signal intensity can weaken, and detector can be aging, and transmission belt also can wear and tear.If do not notice that maintenance is clean, belt surface also has pollutant.This all can affect the terminal signaling that detector receives.Once the deviation outside detector signal generation tolerance, original detected parameters is just no longer applicable, and the detection perform of equipment will be caused to be deteriorated.

For the process of this problem, be that requirement client makes regular check on equipment at present, if find that signal has larger variation, then must investigate the problem of the source of penetrating, detector or belt one by one.The defect done like this is: one, and client does not often accomplish to make regular check on equipment; Two, the evaluation of signal can only depend on sense organ, does not have the quantitative criteria of science.

Summary of the invention

The invention provides a kind of can self-verifying signal and there is the method for the X-ray detection equipment signal evaluation of the quantitative criteria of science, solve the problem due to the equipment Inspection degradation that equipment of making regular check on causes manually cannot be accomplished.

The present invention solves the problems of the technologies described above by the following technical programs: a kind of X-ray machine signal sampling and evaluation method, comprise the steps:

Before step one, equipment dispatch from the factory, ensure that equipment performance has reached one and stablized normal state, then, under specific device parameter, sampled detector signal, is saved as standard signal storehouse;

After step 2, equipment are brought into use, periodic sampling detector terminal signal, during signal sampling, necessarily requires device parameter when equipment and Criterion signal library consistent;

Step 3, signal evaluation, contrast the standard signal in the detector terminal sample of signal of collection and standard signal storehouse, whether in the reasonable scope to evaluate current signal value.

As the technical scheme optimized, this X-ray machine signal sampling and evaluation method also comprise the steps:

Step 4, when the result of signal evaluation be when there is large fluctuation, then alarm is proposed to user.

As the technical scheme optimized, the equipment of described step one dispatches from the factory Criterion signal library, and step 2 equipment brings into use the signal sampling in rear periodic sampling detector terminal signal to comprise following several step:

Step 1: device initialize, by apparatus parameter setting to particular value;

Step 2: multisample is sampled, namely once sampling will extract the sample of the fluctuation range covering signal;

Step 3: calculating mean value, carries out mean value computation to multiple samples of sampling.

As the technical scheme optimized, described specific device parameter comprises penetrates source voltage, electric current, the time shutter of detector and gain.

As concrete technical scheme, when detector is linear array detector, described step 2 is specially to 3: often set up a standard signal or gather a sample signal for signal evaluation, need multiple repairing weld, sampled signal is designated as:

S(i,j)

i＝1,2,3...m

j＝1,2,3...n

M is sampling total amount, and n is pixel number, then the implication of S (i, j) is: the gray-scale value of pixel j during i-th sampling, for arbitrary certain pixel j, it has m sample value, asks for average, that is:

$P\left(j\right)=\underset{i=0}{\overset{m}{\Σ}}S(i,j)$

Then for linear array detector, the signal obtained after a sampling processing is an one-dimensional vector, and it is of a size of pixel number.

As the technical scheme optimized, described step 3 specifically comprises:

Step 1, calculating signal fluctuation vector, specifically calculate absolute change amount absD (i) and relative variation relD (i) of the sampled signal after start use and standard signal;

Step 2, setting evaluation logic, as long as meet arbitrary condition below, then decision signal is abnormal:

1) any one absD (i) is greater than a larger threshold value absT1;

2) have at least p absD (i) to be greater than a less threshold value absT2, p value can be established;

3) any one relD (i) is greater than a larger threshold value relT1;

4) have at least q relD (i) to be greater than a less threshold value relT2, q value can be established;

Step 3, draw evaluation conclusion.

As the technical scheme optimized, when Criterion signal, set up multiple standard reference signal, when standard signal more than one, then establish and count k or proportion threshold value s one by one, be namely all normal compared with at least k standard signal, then think that current demand signal is normal, or the ratio that comparison normal standard signal number accounts for total standard signal number is greater than s, then think that signal is normal.

As concrete technical scheme, for linear array detector, after sampling, its signal value is designated as:

P(i),i＝1,2,3...n

N is the pixel number of detector, and standard signal is designated as: R (i), i=1,2,3...n

Then calculate two proper vectors:

Absolute change amount: absD (i)=| P (i)-R (i) |, i=1,2,3...m

Relative variation: $relD\left(i\right)=\frac{absD\left(i\right)}{P\left(i\right)},i=1,2,3...m.$

As the technical scheme optimized, described step 3 according to the different qualities of signal fluctuation, can propose to user the reason causing fluctuation.

As the technical scheme optimized, if signal fluctuation has globality, namely overall fluctuation amplitude is a close level, then source of penetrating there occurs more serious aging, or detector comprises the reason of its rear end Acquisition Circuit or belt;

If signal fluctuation is only in subrange, in units of sensor assembly, and often clear-cut, be then a certain piece of sensor or comprise its rear end Acquisition Circuit and have exception, if there is the larger fluctuation of the picture element signal of extremely indivedual dispersion, this same instruction is the problem of detector;

If signal fluctuation occurs in random position, having scattered feature, do not have obvious boundary, is then that belt deterioration is excessive, or has more material or pollutant to be detained above, must give replacing or the clean and maintenance of belt.

The present invention has the following advantages compared to existing technology: the signal that can receive at the rear self-verifying detector of each start, and there is the quantitative criteria of science, by the variation discovering device problem timely of signal, solve the problem due to the equipment Inspection degradation that equipment of making regular check on causes cannot be accomplished.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of the method for a kind of X-ray detection equipment signal evaluation that the embodiment of the present invention provides.

Fig. 2 is the process flow diagram of the signal sampling that the embodiment of the present invention provides.

Fig. 3 is a kind of evaluation logic example of the method for the X-ray detection equipment signal evaluation that the embodiment of the present invention provides.

Fig. 4 is the X-ray detection equipment signal wire illustrated embodiment that the embodiment of the present invention provides.

Fig. 5 is the another kind of X-ray detection equipment signal wire illustrated embodiment that the embodiment of the present invention provides.

Embodiment

Elaborate to embodiments of the invention below, the present embodiment is implemented under premised on technical solution of the present invention, give detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

The present invention proposes a kind of X-ray machine signal sampling and evaluation method, helps the undulatory property of user's automatic evaluation complete machine signal.The method: under identical device parameter, regularly (or when at every turn starting shooting) sampled detector signal automatically, is stored into signal library.Adopt certain criterion, the signal that regularly current demand signal and equipment dispatched from the factory is compared, thus assesses it.If signal fluctuation is greater than the tolerance of this equipment, then give the alarm to user.

Refer to shown in Fig. 1, X-ray machine signal sampling of the present invention and evaluation method comprise the steps:

Before step one, equipment dispatch from the factory, Criterion signal library:

Whether signal fluctuation, and will compare with a standard value, correct standard value, in conjunction with the signal evaluation method of science, just can guarantee the rationality of signal evaluation result, therefore, Criterion signal is extremely important.

Before equipment dispatches from the factory, ensure that equipment performance has reached one and stablized normal state, then, (comprise at specific device parameter and penetrate source voltage, electric current, the time shutter of detector and gain) under, sampled detector signal, is saved as standard signal storehouse;

After step 2, equipment are brought into use, periodic sampling detector terminal signal:

After equipment is brought into use, want regularly to sample to detector terminal signal or start just sampling at every turn, all automatically carry out, avoid the uncertainty that artificial sample can be forgotten like this, during signal sampling, necessarily require equipment consistent with device parameter during Criterion signal library, namely identical source of penetrating voltage, electric current, the time shutter of detector and gain, these all affect signal value, along with the continuous use of equipment, even if under identical parameter, signal also can change, as long as but change is in a certain degree, namely Detection results is not affected, be exactly sustainable,

Step 3, signal evaluation:

Standard signal in the detector terminal sample of signal of collection and standard signal storehouse is contrasted, whether in the reasonable scope to evaluate current signal value.This evaluation depends on a set of scientific and effective interpretational criteria, rational judge is made to signal, it is the key of this signal evaluation method, in this evaluation method, antipode feature between sample signal and standard signal should be calculated, also their relative different feature will be calculated, then concrete evaluation logic is designed, as being greater than certain threshold value when antipode and relative different simultaneously, or be greater than certain threshold value separately, or by other judge rule, go to differentiate whether current demand signal larger fluctuation occurs.

As optional step, X-ray machine signal sampling of the present invention and evaluation method also comprise step 4, reminding user:

If in step 3, when the standard signal in the detector terminal sample of signal gathered and standard signal storehouse contrasts, the result passed judgment on is when there is large fluctuation, then to user, alarm to be proposed, different wave characteristics, different component problem may be reacted, cause by source of penetrating is aging as the fluctuation of globality, the fluctuation (in units of detector width) of one piece one piece of locality then causes by detector is abnormal, and the discrete fluctuation of randomness is many because of transport tape loss or be infected with foreign matter and cause.

In above-mentioned signal evaluation method, signal sampling and signal evaluation are two most important modules.Elaborate respectively below.

One, signal sampling

Equipment dispatches from the factory Criterion signal library, and equipment brings into use rear periodic sampling detector terminal signal to evaluate, and all relates to signal sampling, and as shown in Figure 2, signal sampling is divided into following several step:

Step 1: device initialize, by apparatus parameter setting to particular value, device parameter comprises penetrates source voltage, electric current, the time shutter of detector and gain parameter, dispatch from the factory Criterion signal library and equipment of equipment brings into use the parameter of rear periodic sampling detector terminal signal sets should be identical;

Step 2: multisample is sampled, this is to ensure the science of sampling, namely once sampling will extract abundant sample, should how much just calculate enough on earth? this depends on industry knowwhy and the experience of persons skilled in the art, is as the criterion with the fluctuation range roughly covering signal;

Step 3: calculating mean value, carries out mean value computation to multiple sample.

For linear array detector, often set up a standard signal or gather a sample signal for signal evaluation, need multiple repairing weld, sampled signal is designated as:

S(i,j)

i＝1,2,3...m

j＝1,2,3...n

M is sampling total amount, and n is pixel number, then the implication of S (i, j) is: the gray-scale value of pixel j during i-th sampling.

For arbitrary certain pixel j, it has m sample value, asks for average, that is:

$P\left(j\right)=\underset{i=0}{\overset{m}{\Σ}}S(i,j)$

Then for linear array detector, the signal obtained after a sampling processing is an one-dimensional vector, and it is of a size of pixel number.

If planar array detector, then once the rear signal preserved of sampling is a bivector, and it is of a size of horizontal pixel number × longitudinal pixel number.

Certainly also other form can be equivalent to, as adjacent several pixel gray scale is merged into a value.Data buffer storage can be reduced like this.But when signal evaluation, Evaluation accuracy can decline to some extent.

Proposing, when Criterion signal, in order to ensure science further, multiple standard reference signal can be set up.Between these standard signals, difference should be very little.In sampling below, standard signal that can be all with this is compared, pre-conditioned when all meeting certain, or meets pre-conditioned ratio and be greater than certain parameter, then can think that institute's sampled signal is in reasonable fluctuation range.

Two, signal evaluation

As shown in Figure 3, this signal evaluation specifically comprises the following steps the schematic flow sheet of signal evaluation:

Step 1, calculating signal fluctuation vector

For linear array detector, after certain sampling, its signal value is designated as:

P(i),i＝1,2,3...n

N is the pixel number of detector, and standard signal is designated as: R (i), i=1,2,3...n

Then calculate two proper vectors:

Absolute change amount: absD (i)=| P (i)-R (i) |, i=1,2,3...m

Relative variation: $relD\left(i\right)=\frac{absD\left(i\right)}{P\left(i\right)},i=1,2,3...m$

Step 2, setting evaluation logic

As long as meet arbitrary condition below, then decision signal is abnormal:

1) any one absD (i) is greater than a larger threshold value absT1;

2) have at least p absD (i) to be greater than a less threshold value absT2, p value can be established;

3) any one relD (i) is greater than a larger threshold value relT1;

4) have at least q relD (i) to be greater than a less threshold value relT2, q value can be established.

Certainly, sometimes as required, the decision condition of abnormal signal can be revised, as being revised as: 1), 2) for or relation, 3), 4) be or relation, 1), 2) with 3), 4) be and relation, namely 1) with 2) meet one, and 3) with 4) satisfied one.

The above-mentioned comparison for judging present sample signal and a standard signal, if more than one of standard signal, then can establish and count k or proportion threshold value s one by one, namely be all normal compared with at least k standard signal, then think that current demand signal is normal, or the ratio that comparison normal standard signal number accounts for total standard signal number is greater than s, then think that signal is normal.

Step 3, draw evaluation conclusion

Through the analysis of evaluation logic, learn when whether normal time signal is, if normally, then equipment can enter testing process, if abnormal, then system can propose alarm or prompting to user, according to the different qualities of signal fluctuation, to user, the possible reason causing fluctuation is proposed.

(source is penetrated and detector position is all normal in the completely normal situation of equipment, ray can not be there is be blocked, detector does not receive the situation of signal), if signal fluctuation has globality, namely overall fluctuation amplitude is a close level, then be likely that the source of penetrating there occurs more serious aging, but also not exclusively getting rid of is the reason that detector comprises its rear end Acquisition Circuit or belt, because the change of penetrating source signal is overall, detector is spliced by one piece of one piece of sensor, each sensor signal there are differences, its fluctuation is also often different, and the breakage of belt or on it position of retentate be generally random, its signal fluctuation of causing is often neither entirety, neither in units of detector, only arbitrarily be distributed in everywhere.

If signal fluctuation is only in subrange, be one piece one piece, in units of sensor assembly, (be not sensor pixel, a sensor assembly is generally made up of the pixel of some.The sensor assembly of varying number then forms the detector of different detection length.), and often clear-cut, then very large may be that a certain piece of sensor (containing its rear end Acquisition Circuit) has exception, certainly, the larger fluctuation of the picture element signal of extremely indivedual dispersion also may occur sometimes, this equally instruction be the problem of detector.Refer to Fig. 4, transverse axis-pixel in Fig. 4, the longitudinal axis-gray-scale value, the detector in Fig. 4 is made up of 8 pieces of sensor assemblies, and the line being parallel to the longitudinal axis is module edge signal wire, is the separator bar of disparate modules.Because multirow sample mean, the signal gray-scale value in Fig. 4 is not integer.Can see, the different sensors module by signal property of there are differences, the 1st in Fig. 4, the signal fluctuation of 5,7 modules is comparatively obvious.In reality, the signal amplitude of different sensors module may have larger difference.Fig. 4 is only signal.Comparison diagram 5 and Fig. 4, first piece of sensor signal amplitude occurs obviously to move down, then remind this block sensor (or its rear end Acquisition Circuit) abnormal.Notice that the displaying ratio of Fig. 5 with Fig. 4 is different.

If signal fluctuation occurs in random position, have scattered feature, do not have obvious boundary, instruction may be that belt deterioration is excessive, or has more material or pollutant to be detained above, must give replacing or the clean and maintenance of belt.

In this patent, we not only propose a kind of signal evaluation method, for helping the fluctuation of client's automatic analysis equipment signal, also refer to the concept of signal library.

Signal library contains all standard signals and equipment sampled signal in use, signal library mainly helps client to check the overall signal variation tendency of equipment run duration, estimating signal decay rate, having directive significance to understanding source of penetrating, the performance of detector and belt and life-span.Due to the otherness of every platform equipment, the signal library content of multiple devices then provides the stability information about this model device.This Optimal improvements for equipment is without benefits.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.

Claims (10)

1. X-ray machine signal sampling and an evaluation method, is characterized in that: comprise the steps:

Before step one, equipment dispatch from the factory, ensure that equipment performance has reached one and stablized normal state, then, under specific device parameter, sampled detector signal, is saved as standard signal storehouse;

After step 2, equipment are brought into use, periodic sampling detector terminal signal, during signal sampling, necessarily requires device parameter when equipment and Criterion signal library consistent;

Step 3, signal evaluation, contrast the standard signal in the detector terminal sample of signal of collection and standard signal storehouse, whether in the reasonable scope to evaluate current signal value.

2. a kind of X-ray machine signal sampling as claimed in claim 1 and evaluation method, is characterized in that: also comprise the steps: after described step 3

Step 4, when the result of signal evaluation be when there is large fluctuation, then alarm is proposed to user.

3. a kind of X-ray machine signal sampling as claimed in claim 1 and evaluation method, is characterized in that:

The equipment of described step one dispatches from the factory Criterion signal library, and step 2 equipment brings into use the signal sampling in rear periodic sampling detector terminal signal to comprise following several step:

Step 1: device initialize, by apparatus parameter setting to particular value;

Step 2: multisample is sampled, namely once sampling will extract the sample of the fluctuation range covering signal;

Step 3: calculating mean value, carries out mean value computation to multiple samples of sampling.

4. a kind of X-ray machine signal sampling as described in any one of claims 1 to 3 and evaluation method, is characterized in that: described specific device parameter comprises penetrates source voltage, electric current, the time shutter of detector and gain.

5. a kind of X-ray machine signal sampling as claimed in claim 3 and evaluation method, it is characterized in that: when detector is linear array detector, described step 2 is specially to 3: often set up a standard signal or gather a sample signal for signal evaluation, need multiple repairing weld, sampled signal is designated as:

S(i,j)

i＝1,2,3...m

j＝1,2,3...n

M is sampling total amount, and n is pixel number, then the implication of S (i, j) is: the gray-scale value of pixel j during i-th sampling, for arbitrary certain pixel j, it has m sample value, asks for average, that is:

Then for linear array detector, the signal obtained after a sampling processing is an one-dimensional vector, and it is of a size of pixel number.

6. a kind of X-ray machine signal sampling as claimed in claim 1 and evaluation method, is characterized in that: described step 3 comprises:

Step 1, calculating signal fluctuation vector, specifically calculate absolute change amount absD (i) and relative variation relD (i) of the sampled signal after start use and standard signal;

Step 2, setting evaluation logic, as long as meet arbitrary condition below, then decision signal is abnormal:

1) any one absD (i) is greater than a larger threshold value absT1;

2) have at least p absD (i) to be greater than a less threshold value absT2, p value can be established;

3) any one relD (i) is greater than a larger threshold value relT1;

4) have at least q relD (i) to be greater than a less threshold value relT2, q value can be established;

Step 3, draw evaluation conclusion.

7. a kind of X-ray machine signal sampling as claimed in claim 6 and evaluation method, it is characterized in that: when Criterion signal, set up multiple standard reference signal, when standard signal more than one, then establish and count k or proportion threshold value s one by one, be namely all normal compared with at least k standard signal, then think that current demand signal is normal, or the ratio that comparison normal standard signal number accounts for total standard signal number is greater than s, then think that signal is normal.

8. a kind of X-ray machine signal sampling as claimed in claim 6 and evaluation method, is characterized in that: for linear array detector, and after sampling, its signal value is designated as:

P(i),i＝1,2,3...n

N is the pixel number of detector, and standard signal is designated as: R (i), i=1,2,3...n

Then calculate two proper vectors:

Absolute change amount: absD (i)=| P (i) R (i) |, i=1,2,3...m

Relative variation: $relD\left(i\right)=\frac{absD\left(i\right)}{P\left(i\right)},i=1,2,3...m.$

9. a kind of X-ray machine signal sampling as claimed in claim 6 and evaluation method, is characterized in that: described step 3 according to the different qualities of signal fluctuation, can propose to user the reason causing fluctuation.

10. a kind of X-ray machine signal sampling as claimed in claim 9 and evaluation method, it is characterized in that: if signal fluctuation has globality, namely overall fluctuation amplitude is a close level, source of then penetrating there occurs more serious aging, or detector comprises the reason of its rear end Acquisition Circuit or belt;

If signal fluctuation is only in subrange, in units of sensor assembly, be then a certain piece of sensor or comprise its rear end Acquisition Circuit and have exception, if there is the larger fluctuation of the picture element signal of extremely indivedual dispersion, this equally instruction be the problem of detector;

If signal fluctuation occurs in random position, having scattered feature, do not have obvious boundary, is then that belt deterioration is excessive, or has more material or pollutant to be detained above, must give replacing or the clean and maintenance of belt.