JPH02110358A - Automatic selecting method of vegetable - Google Patents

Abstract

PURPOSE:To distinguish a recessed part of a body from a void inside said body to be checked by limiting a take-out region of a magnetic resonance signal in said body to be checked to a desired size in directions of two gradient magnetic fields, taking out and processing a magnetic resonance signal from said limited region, whereby the inner structure of said body to be checked is checked. CONSTITUTION:Main components of a static magnetic field generating unit 4 are permanent magnets 5a, 5b which add a predetermined static magnetic field to a body to be checked, for example, a water melon 3. A first selecting wireless frequency pulse 16 forming a high frequency magnetic field is added to the body 3 in existence of a first gradiant magnetic field Gx, and a succeeding second selecting wireless frequency pulse 17 is added to the body 3 in existence of a gradient magnetic field Gy orthogonal to the first gradient magnetic field Gx, so that the take-out region of a magnetic resonance signal in the body 3 is limited to an optional size in the directions of the two gradient magnetic fields Gx and Gy. A magnetic resonance signal is taken out from this limited region and processed to check the inner structure of the body 3. Accordingly, data of a recessed part on the surface of the body 3 is removed from the magnetic resonance signal, whereby the presence or absence of a void inside the body 3 can be correctly detected.

Description

[Detailed description of the invention] (b) Industrial application field The present invention relates to an automatic sorting method for vegetables and fruits.

In particular, the present invention relates to an automatic sorting method for nondestructively inspecting moisture content and detecting internal cavities using the magnetic resonance phenomenon of atoms constituting fruits and vegetables.

(b) Prior Art In recent years, automatic sorting devices for vegetables, fruits, etc. are becoming popular in agricultural cooperatives (hereinafter referred to as ATA) in these production areas. This is because: - By quantifying the criteria for grading, it becomes possible to sort according to the specified standard values, increasing trust in the market.

■ Selecting high-quality products that sell at relatively high prices,
We can adjust and ship the grade to match the destination's requests, etc.
It is effective in sales strategies that take advantage of the characteristics of the production area.

■ Labor-saving can be achieved in grade sorting, which requires the most manpower.

Due to reasons such as.

For example, in an automatic watermelon sorting system, a conventional device that uses a combination machine to analyze the shock waves generated by the tapping sound of a professional inspector and determines whether there are internal tears is agribusiness°88.
Vol, 3 NQIOPP.

11-17 Introduced in System Agriculture Co., Ltd., a device that calculates and determines the area of a region corresponding to a cavity by taking a fluoroscopic image using electromagnetic waves such as X-rays and binarizing it using a computer. was introduced in the Agricultural News dated September 26, 1986.

In the former case, the existing cavity cannot be accurately captured depending on the location of the tap, and it may cause damage to the subject. In the latter case, the inside can be seen as an image without destroying it, so it is difficult to inspect. Although the accuracy can be significantly improved, since it is a projection image, sufficient contrast resolution cannot be obtained when dealing with objects with different external shapes, and it is not suitable for moving objects. The disadvantage of this is that it gives a bad impression in the market.

(c) Problems to be Solved by the Invention Since the signal of the object to be examined by magnetic resonance is processed and used at each cross section, it is not suitable as a method for sorting fruits and vegetables of this kind. For this reason, it is necessary to extract signals from a specific cross-section of the object to reduce data processing time, but with this method, there is a risk of misidentifying a depression on the object's surface as an internal cavity. . The problem to be solved by the present invention was made in consideration of such problems, and it is a method to remove the data of the depressions on the surface of the specimen from the magnetic resonance signal and accurately detect only the presence or absence of internal cavities. It's about finding out.

(d) Means for solving the problem In an automatic fruit and vegetable sorting method in which grades are sorted by inspecting the internal structure of a gradient magnetic field on a subject placed in a static magnetic field generator, the high frequency magnetic field is applying a pulse to the subject in the presence of a gradient magnetic field orthogonal to the first gradient magnetic field to limit a magnetic resonance signal extraction region in the subject to an arbitrary size with respect to the two gradient magnetic field directions; Magnetic resonance signals are extracted from this limited area and processed to examine the internal structure of the subject.

By applying magnetic resonance signals only from a desired region of an arbitrary cross section of the object, unnecessary data on the surface of the object is removed.

(f) Example The automatic fruit and vegetable sorting method of the present invention will be described in detail below with reference to an example of the drawings.

Figure 2 shows the configuration of the automatic sorting machine, and in the figure (1)
is the fruit delivery conveyor, (2) is the watermelon as the test object (3)
A special tray for placing the on the conveyor (1), (4)
is a static magnetic field generating device which is the heart of the automatic sorting machine, and the watermelon (3) carried by the conveyor (1) passes through the static magnetic field generating device (4). The main component of the static magnetic field generator (4) is that the entire static magnetic field generator (4) is radio-shielded to take measures against radio interference caused by external noise when applying a high-frequency magnetic field. is a magnet that applies a constant static magnetic field to the watermelon (3), especially a permanent magnet (5a) (5bl) that is easy to maintain.Watermelon (3) has a rich moisture content and the obtained magnetic resonance signal is strong. Permanent magnets (5a) (5b) are the best since they can be used in a relatively low magnetic field, and from the viewpoint of maintenance costs, etc. (6a)
(6b) is a flat yoke located vertically at intervals;
(7a) (7b) is the upper and lower yoke (6a) (6b)
These yokes (6a) (6b) are four columnar yokes that maintain a distance of l and form a closed magnetic circuit.
, (7al (7b) are both made of iron, and the permanent magnet (5a) (5bl is attracted to the inner surface of the upper and lower yokes (6a) and (6b) by their own magnetic force. Also, the permanent magnet (5a) (5b) has an iron pole piece magnetically attracted to each opposing surface side of the permanent magnets (5a) (5b), and the permanent magnets (5a) (5b) are magnetized by the pole pieces.
) increases the uniformity of the static magnetic field created. In addition, although the permanent magnet (5al (5b)) has temperature dependence, it itself has a large heat capacity, so a temperature control device is required to maintain a constant average temperature.For example, an air conditioner with a temperature controller is required. It is preferable to install a magnetic powder etc. into the static magnetic field generator (11), because it will be attracted to the parts constituting each of the magnetic circuits. Reception = 1 Radio frequency (hereinafter referred to as RF) transmission and reception that amplifies and detects the signal received at Il (9)
f1. (+21 is a pulse programmer that sends preprogrammed control signals to the RF transmission/reception TFI (11) and the power supply (10); (13) is this pulse programmer (
12) may be provided with a conductive curtain to prevent data or frame failure, or a cylindrical shield may be provided to increase the penetration distance to the center of the static magnetic field generator (4).

On the surface of the pole piece, gradient magnetic field generating coils (
8a) (8b) is fixed, and a receiving coil (9) that also serves as a high-frequency magnetic field pulse coil is provided in the vicinity of the watermelon (3) and has a portion where the watermelon (3) and the conveyor belt (1a) pass through. .

(lO) is the gradient magnetic field generating coil (8al (8bl
A power source (11) is a device that supplies high frequency power to the high frequency magnetic field generating coil (9).

Depending on the number of watermelons to be discriminated, one to three automatic sorting machines having the static magnetic field generator (4) having the above configuration as a main body are installed together with the conveyance line of the fruit conveyor (1).

By the way, the best way to obtain the magnetic resonance signal of watermelon (3) is
A simple method is to stop the conveyor (1) by -1, position the watermelon (3) at the center of the static magnetic field generator (4), and with the watermelon (3) stationary, move the gradient magnetic field generator coil ( 8a) (8b) and high frequency magnetic field generation coil (9
) with a predetermined pulse sequence (one sequence per piece as described later), a gradient magnetic field is superimposed on the static magnetic field, and a high-frequency magnetic field pulse is applied to the watermelon (3). Some or all of the water protons are excited, and a gradient magnetic field is applied to read out the magnetic resonance signal of the excited water protons (when one signal is Fourier transformed and decomposed into a spectrum, it can be divided into various information). while detecting. The signal obtained in this way is RF
Through the transceiver (11), the signals are amplified, detected, digitized, and sent to the central information processing unit, where the data is processed.It is a device that can measure the proton density of watermelon (3) and its layer, and is used for nuclear magnetic resonance medicine. By avoiding the steps described in detail in NMR Medicine - Basic and Clinical - edited by Ishiha Research Association, the normally required 256 or 128 pulse sequences can be completed in one time, increasing the discrimination speed to 1 pulse/second. That's what I do.

The obtained distribution data is a partial (point, line, area, volume) signal intensity distribution of watermelon (3), and if there is a cavity in a part of the observation area, the signal will be zero only there, so there is no cavity. If you can determine the presence or absence of , you can determine if it is true.

Of course, the above discrimination is not limited to watermelons, but can also be applied to automatic sorting of various fruits and vegetables such as cabbage and apples.

However, there is one problem with the above method. That is, using the spin echo method as shown in Figure 3, 90' and 180'
When an RF excitation pulse is applied to a watermelon (3) in the presence of gradient magnetic fields Gy and G2, the spectrum obtained by Fourier transformation of the echo signal (magnetic resonance) obtained with one pulse sequence is shown in Figure 11. (3) The internal cavity (14) and the surface depression (15) are indistinguishable. 4th LK (a), (b), and (c) show a good watermelon, a watermelon with a cavity (14), a watermelon with a dent (15), and its svector, but the cavity (14)
and the concavity (15) appear in exactly the same shape in the spectrum.

A pulse sequence for solving this problem is shown in FIG. The part on the right side of the broken line This is for extracting magnetic resonance signals from an arbitrary XY plane.

On the other hand, the part to the left of X-X is a sequence consisting of gradient magnetic fields Gx and Gy perpendicular to the two directions for selective excitation and two 90° RF excitation pulses. The excitation process in this pulse sequence is shown in Figure 5. Shown below. First, the gradient magnetic field G
When a watermelon (3) is selectively excited in the X direction by x and the first 90° RF excitation pulse (16), the proton nuclear spin in the shaded area in the same figure (a) is tilted by 90° in the XY thousand plane. It will be done. Next, the gradient magnetic field Gy and the second 90'
When the watermelon (3) is selectively excited in the X direction by the RF excitation pulse (17), the part excited by the gradient magnetic field Gx and the 190゛ excitation pulse (16), as indicated by diagonal lines in FIG. The spins in the intersecting region are further inverted by 90 degrees, meaning they are inverted by 180 degrees, and the spins in other parts are inverted by 90 degrees.

After this, when the pulse sequence on the right side of X-X is executed, an echo signal is obtained only from the part where the spins are tilted 180 degrees by the pulse sequence on the left side.
It will not come out from the part where it has been knocked down by 90 degrees. Using this method, we extract the magnetic resonance signals for the three watermelons shown in Figures 4(a), (b), and (c), Fourier transform them, and obtain spectra as shown in Figures 6(a), (b). ) As shown in (c), the non-defective watermelon and the watermelon with a dent (15) on the surface have the same rectangular spectrum, and only the watermelon (3) with a cavity (!4) is distinguished from the two. The difference between good watermelons and watermelons with dents (15) on the surface can be determined visually or by using another automatic sorting machine.
Discrimination can be made by obtaining magnetic resonance signals only from the 0° tilted portion.

Figure 7 shows magnetic resonance from different XY slice planes by excitation multiple times while sequentially changing the 90' RF excitation pulse frequency on the right side of X-X in the pulse sequence of Figure 1. This is a bullshit showing a multi-slice technique that can also obtain signals. In this way, it is also possible to obtain magnetic resonance signals from different cross sections of the watermelon (3) and obtain three-dimensional internal information of the watermelon (3) from this magnetic resonance. However, in this case, it is better to keep the number of slice planes small in order to shorten the inspection time as much as possible.

Furthermore, by adjusting the gradient magnetic field^ and the frequency of the RF selective excitation pulse in the pulse sequence shown in Figure 1,
It is possible to enlarge or reduce the area where magnetic resonance signals are emitted in accordance with the size of the subject.

(G) Effects of the Invention As described above, the present invention applies a gradient magnetic field and a high-frequency magnetic field to a subject placed in a static magnetic field generator to cause magnetic resonance, and processes signals caused by this magnetic resonance. In addition to the automatic fruit and vegetable sorting method in which grades are sorted by inspecting the internal structure of the specimen, the first step is to form the high-frequency magnetic field. ! applying a selective radio frequency pulse to the subject in the presence of a first gradient magnetic field, followed by a second selective radio frequency pulse limiting the first region to an arbitrary magnitude with respect to the two gradient magnetic field directions; By extracting the magnetic resonance signal from this limited area and processing this signal to inspect the internal structure of the specimen, it is possible to clearly distinguish between the recesses on the surface of the specimen and the internal cavities, which can be used to improve grade selection. It has the effect of improving the degree of

[Brief explanation of the drawing]

FIG. 1 is a front view showing an outline of the method used in the automatic fruit and vegetable sorting method of the present invention, FIG. 3 is a diagram of a pulse sequence without the pulse sequence for determining the signal extraction area shown in FIG. 1, and FIG. b) and (c) are spectrum diagrams of watermelons of good quality, with cavities, and with surface depressions, respectively, obtained using the pulse sequence shown in Fig. 3, and Fig. 5 (a) and (b) are the signal extraction shown in Fig. 1. 6(a), (b), and (c) are spectrum diagrams of a watermelon corresponding to FIG. 4 obtained using the pulse sequence of FIG. 1; FIG. 7 shows a signal spectrum diagram emitted from each cross section by a multi-slice method using the pulse sequence of FIG. 1 as a basic form. (4)... Static magnetic S generator, (3)... Subject (watermelon), (GX) (GF) (Gxl-... Gradient magnetic field, <16) (17+=-Radio frequency pulse.

Claims (1)

[Claims] (1) Applying a gradient magnetic field and a high-frequency magnetic field to a subject placed in a static magnetic field generator to cause magnetic resonance, and processing the signals caused by this magnetic resonance to inspect the internal structure of the subject. In an automatic fruit and vegetable sorting method for sorting grades, a first selective radio frequency pulse forming the high frequency magnetic field is applied to the subject in the presence of a first gradient magnetic field, followed by a second selective radio frequency pulse forming the high frequency magnetic field.
A selected radio frequency pulse is applied to the subject in the presence of a gradient magnetic field orthogonal to the first gradient magnetic field, so that the magnetic resonance signal extraction area in the subject is set to an arbitrary size with respect to the two gradient magnetic field directions. 1. An automatic fruit and vegetable sorting method characterized by extracting a magnetic resonance signal from this limited area and processing this signal to inspect the internal structure of a subject.