JPS58166278A - Separating method of time series signal - Google Patents

Abstract

PURPOSE:To improve range resolution in a pulse reflecting method using a high sensitivity transmitter and receiver whose Q is comparatively large, by convolution-operating a received echo signal, and separating each echo signal. CONSTITUTION:In response to a pulse input signal (f) from a signal generator 1, an impulse (h) is outputted from a transmitter 4, and a reflected pulse reflected by a reflecting body 6 is received by a receiver 5. This received signal is convolve-operated with an impulse response (g) of a convolver 2 and obtains a resulted signal (d) by a separated echo signal. In this case, when the response (g) is decided so that a square-law error of the signal (d) and a desired waveform becomes minimum, range resolution is improved in a pulse reflecting method using a high sensitivity transmitter and receiver whose Q is comparatively large, without making Q small.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for improving distance resolution in an ultrasonic tomography apparatus using a pulse reflection method.

One method for this is to minimize the Q of the transducer and generate a high Lqi acoustic pulse.This method is commonly used5. If Q is reduced using this method, it is certainly possible to generate odor acoustic pulses, but on the other hand, the electric shock of the transducer is reduced.

This invention is a pulse reflection method using a transducer with a comparatively large Q and a relatively strange shape.As a method for improving distance resolution when multiple reflected waves overlap, the received signal is processed by convolution calculation, and each reflected wave is This method achieves the desired purpose by separating the waves.

Hereinafter, a method of designing a processing system for separating overlapping reflected waves that occurs when a transducer with a large Q according to the present invention is used will be described.

Figure 1 shows a basic system using the pulse reflection method. As shown in FIG. 1, the input signal is f, the received signal is y1, and the impulse response of the acoustic system is h. Further, let g be the impulse response of the processing system, that is, the convolver, and let d be the signal finally obtained.・In other words, the signal generator 1 generates a pulse wave input signal f, sends it to the transmitter 4, and obtains the impulse response of the acoustic system of the reflector 6, such as the growth rings of a tree, the constituent layers of the human body, etc. , this is given to the convolver 2 as a received signal y from the receiver 5, and this and the impulse response g of the convolver are
The result signal d obtained by convolving and processing is displayed on the display 3.
This figure shows an example of a configuration in which a display image is obtained by applying 1'f to 1' to obtain a display image.

In this case, d is given by the following equation. J d=f*h*g=FHg ・・・・・・・・・・・・
・・・・・・・・・(1) However, * indicates convolution,
F and H are the matrices that perform the calculation, and are given by the following equation.Here, assuming that K is the desired waveform that can improve the distance resolution when there is only one reflected wave, and the squared error of d and K is δ, then , δ is given by the following equation.

δ=(d-K) (d-K)t =gtHtFtFHg-2gtHtF'+KtK
......(2) However, t on the right shoulder is 1 indicating the transposition of a matrix, etc. From the above, δ in equation (2) is a received signal in which multiple reflected waves overlap, and g Distance resolution can be improved by convolving and separating each reflected wave. Also g
is calculated using the following formula.

HtFtFHg=HtFtK ・・・・・・
・・・・・・・・・・・・(3) In this way, the input signal f
By determining the impulse response of the acoustic system and assuming the desired waveform K, the processing system g can be easily designed from equation (3).

In addition, this type of processing system uses a delay element such as a BBD.
This can be easily achieved with an IR filter.

The calculation was carried out by setting f to be a signal obtained by gated a sine wave with three wavelengths, setting Q of the transducer to 10, setting the number of stages of g to 64, and setting K to a signal obtained by gated a sine wave by one wavelength.

First, the minimum square error δ was calculated when the desired waveform K is further delayed by N sample time after the propagation of the acoustic system. The results are shown in FIG. From this figure, it can be seen that δ decreases as N increases and becomes minimum when N is 56. G at this time again
is shown in Figure 3.

Next, we will consider separating a received signal in which four reflected waves overlap. Assume that the received signal has a waveform in which the amplitude ratio of each reflected wave is l: o, s: 0.7: t, and the intervals between each reflected wave are 50, 20, and 30, and the waveform at this time is As shown in the figure.

Generally, when the interval between reflected waves is short, overlapping occurs as shown in FIG. 4, and as a result, the distance resolution of the resulting reflected image is reduced.

Further, FIG. 5 shows the result of convolving the signal in FIG. 4 with g in FIG. 3. From this result, four reflected waves are separated,
It can be seen that the amplitude ratio is also preserved. However, as previously explained, each separate reflected wave appears delayed by 56 sample times.

According to the present invention, as described above, even when using a transducer with a relatively large Q and high transmitting/receiving sensitivity, multiple The reflected waves can be separated and the distance resolution can be improved.2 This method is effective as a distance resolution improvement method when the reflected waves are weak and a highly sensitive transducer is required or when the intervals between the reflected waves are short. It is.

[Brief explanation of drawings]

The drawings show examples, and Fig. 1 is a sonic tomography^, a schematic diagram of the main part configuration of the device, Fig. 2 is a line diagram of ・Ien i + Ii 1.2 east error, and Figs. 3 to 5 are diagrams. FIG. 3 shows the response g of the convolver, FIG. 4 shows the received signal h, and FIG. 5 shows the waveform of the processed signal d. DESCRIPTION OF SYMBOLS 1... Signal generator, 2... Convolver, 3... Display device, 4... Wave transmitter, 5... Wave receiver, 6... Reflector. Patent applicant Ya 2 and a half 30 IME 4th opening 5k IME Procedural amendment (spontaneous) September 3, 1980 Director General of the Japan Patent Office 1. Indication of case Patent application 1984-47465 2. Title of the invention Separation method of series signals 3, relationship with the case of the person making the amendment Patent applicant Ma141 2-10-45 Kamitasaki, Tokyo Parts Industry Specification/Drawing 5, Contents of amendment No change)

Claims (1)

[Claims] As a method for improving separation resolution when multiple reflected waves overlap, a time-series signal separation method that processes the received signal using a convolution calculation and separates each reflected wave.