CN107185850B - Corn seed activity detection device based on hyperspectral imaging and electrical impedance - Google Patents

Abstract

The invention discloses a corn seed vitality detection device based on hyperspectral imaging and electrical impedance, comprising: a material box storing corn seeds to be tested, and one or more chutes for outputting single seeds are connected to the bottom of the material box; The conveyor belt is equipped with a detection board, and the detection board is provided with a detection hole for receiving the seeds output by the chute, and each detection hole is provided with an electrode pair; an impedance analyzer connected to the electrode pair is used to apply an excitation current to the seeds through the electrodes , according to the electrical impedance data between the electrode pairs; the hyperspectral imaging device installed above the conveyor belt is used for real-time imaging of the corn seeds in the detection hole; the computer is used for comprehensive analysis of the electrical impedance data and hyperspectral data. The vigor of corn seeds is graded; the collection container at one end of the conveyor belt is provided with at least two holding tanks for accepting corn seeds of different vigor grades. The invention has the advantages of non-destructive, high degree of automation, rapid detection, high screening efficiency and the like.

Description

A corn seed vigor detection device based on hyperspectral imaging and electrical impedance

technical field

The invention relates to the field of vigor detection of plant seeds. Specifically relates to a corn seed vigor detection device based on hyperspectral imaging and electrical impedance.

Background technique

Corn is my country's largest food crop, and the quality of corn seeds is directly related to the benefits of seed companies and farmers. In the annual seed quality disputes, there is often a big gap between the germination rate in the laboratory and the emergence rate in the field. The indoor germination rate meets the label value, but the field emergence rate is difficult to meet the standard. This is the problem caused by the vitality of the seeds. Seed vigor refers to the robustness of seeds, which is the sum of seed germination and emergence rate, seedling growth potential, plant stress resistance and production potential, and is an important indicator of seed quality. Vitality detection of seeds before sowing becomes an essential part and plays a vital role in agricultural production.

Seed vigor assays can be divided into physiological assays and biochemical assays. Physiological assays measure the germination characteristics and growth indicators of seeds; biochemical assays measure special biochemical reactions related to seed vigor, such as enzyme system activity. At present, the traditional methods for measuring the vigor of corn seeds mainly include: cold resistance test method, seedling growth test method, accelerated aging test method, electrical conductivity test method, etc. The above-mentioned traditional seed viability detection methods have disadvantages such as large measurement workload, poor repeatability, long measurement cycle, large environmental impact, and easy damage to seeds.

Hyperspectral imaging technology is a comprehensive technology integrating detector technology, precision optical machinery, weak signal detection, computer technology, and information processing technology. Features such as "unity". In the ultraviolet, visible, near-infrared and mid-infrared regions, hundreds of spectral data with very small band intervals and continuous spectra are obtained. Each pixel in the spectrum corresponds to a continuous spectrum, which can reflect the internal composition information of the object to be measured, and has broad application prospects in the field of agricultural research.

Seeds have the property of absorbing light, and in the case of different beam intensities and beam densities, the surface reflection and body reflection of the seeds are different. The optical properties of the seeds are related to the internal physiological indicators of the seeds. Therefore, hyperspectral imaging technology can be used to non-destructively detect the vigor of the seeds using the optical properties of the seeds.

On the other hand, seed water content is closely related to seed vigor. When the water content is high, the respiration of the seeds is strengthened, the consumption of stored substances in the seeds is accelerated, and the vitality of the seeds decreases. The accumulation of organic matter is less, and the enzyme activity is low, which cannot meet the energy required for seed germination, resulting in a low germination rate. On the other hand, too low a water content will result in too little respiration and reduced seed vigor. Therefore, suitable water content is an important indicator of seed vigor, and the determination of seed water content can also reflect seed vigor to a certain extent.

In order to overcome the shortcomings of existing technologies, the present invention is different from traditional physiological assays and biochemical assays. Through hyperspectral imaging technology and electrical impedance technology, on the one hand, it provides a method for online non-destructive detection and screening of corn seed vigor; on the other hand, through The combination of the two technologies can obtain multi-dimensional information on the vigor of corn seeds, which greatly improves the accuracy of detection. For samples that have obtained optical and electrical information, determine the vigor level of single seeds through germination experiments, and establish a seed optical/electrical/image information-viability model library. Comparing the model library, you can know the current vigor level of corn seeds.

The object of the present invention is to detect the hyperspectral image and electrical impedance of corn seeds through a non-destructive method, combine the corn seed spectrum/image information-vitality model and the electrical impedance-vitality model to obtain multi-dimensional corn seed vitality information, through a device Realize online non-destructive detection and screening of corn seed vigor. For this purpose, this invention provides a device for non-destructive online detection and screening of corn seed vigor using hyperspectral imaging technology and electrical impedance technology.

Contents of the invention

The invention discloses a corn seed vigor detection device based on hyperspectral imaging and electrical impedance, which realizes non-destructive online detection of corn seed vigor based on hyperspectral imaging and electrical impedance, and is used for screening high vigor corn seeds. This system fundamentally solves the disadvantages of low efficiency, high cost, low accuracy, and damage to seeds in traditional seed vigor detection methods. The innovative combination of hyperspectral imaging technology and electrical impedance technology uses the optical and electrical properties of corn seeds to obtain multi-dimensional vitality information of corn seeds. On the one hand, the device and method are provided for the online non-destructive detection of corn seed vigor, and on the other hand, the accuracy is greatly improved. At the same time, the device can also screen the corn seeds in real time through the mechanical device according to the detection results. The device has the advantages of non-destructive detection, high degree of automation, rapid detection, and efficient screening.

For realizing above-mentioned purpose of the invention, the concrete technical scheme that the present invention adopts is as follows:

A corn seed vigor detection device based on hyperspectral imaging and electrical impedance, comprising:

A feed box for storing corn seeds to be tested is connected to the bottom of the feed box with one or more chutes for outputting single seeds;

A conveyor belt equipped with a detection plate, the detection plate is provided with a detection hole for receiving the seeds output by the chute, and each detection hole is provided with an electrode pair;

An impedance analyzer connected to the electrode pair is used to apply an excitation current to the seed through the electrode, according to the electrical impedance data between the electrode pair;

The hyperspectral imaging device installed above the conveyor belt is used for real-time imaging of corn seeds in the detection hole;

Computer for comprehensive analysis of electrical impedance data and hyperspectral data to grade corn seed vigor;

The collection container at one end of the conveyor belt is provided with at least two holding tanks for accepting corn seeds of different vigor levels.

In the present invention, the corn seeds pass through the chute at the bottom of the material box so that the batches of corn seeds to be tested are evenly distributed on the conveyor belt. seed.

For the corn seeds in the detection hole, apply a variety of different suitable excitation currents I to the corn seeds by setting the experimental program, and simultaneously detect the impedance value and phase angle between the electrodes, automatically save and record the data, and through the adjustment of the experimental program, you can Provide excitation currents with different currents, frequencies, and waveforms to meet the needs of different scenarios.

The detection board has a single-row structure and integrates two or more detection holes equipped with electrodes. The size and depth of the hole corresponds to the size of the corn seed, which can fit exactly one corn seed.

Preferably, the electrode pair is two electrodes with the same surface area and interlaced with each other.

Preferably, each electrode includes an arc-shaped plate connected to the electrode wire, the arc-shaped plates of the two electrodes are symmetrically arranged, and the two arc-shaped plates are provided with grid bars extending oppositely and interlaced at intervals.

Preferably, the electrode pair is made of graphite.

In the present invention, the electrodes are two interlaced electrode structures with substantially the same surface area, and are made of graphite, which is beneficial to extracting the impedance information of corn seeds.

Preferably, the hyperspectral imaging device includes two sets of imaging spectrometers with wavelength ranges of 380-1023nm and 874-1734nm respectively.

In the present invention, the hyperspectral imager includes: two sets of imaging spectrometers (380-1023nm and 874-1734nm), CCD cameras, imaging lenses of XNP 1.4/17 (380-1023nm) and OLES15 (874-1734nm), Fiber-Lite DC950 line light source and dark box. The spectral resolution of 380-1023nm is 2.8nm, and the spectral resolution of 874-1734nm is 5nm. The lens height, light source, and conveyor belt movement speed can be adjusted. The computer can perform hyperspectral imaging on corn seeds with different movement speeds at different heights by adjusting the exposure time, light source intensity, frame number of pixel acquisition, and resolution. The needs of different scenarios. Among them, hyperspectral imaging data processing includes data preprocessing, feature extraction and model comparison.

Furthermore, the surface of the conveyor belt is made of black non-reflective material, which can move at different constant speeds.

Preferably, the detection board is a cuboid, and several detection boards are sequentially spliced and fixed on the conveyor belt.

In the present invention, multiple rows of detection boards are installed on the conveyor belt, which can be disassembled by adopting a modular design, and different sizes of detection boards can be selected for different sizes of seeds, which has high universality.

The computer has built-in editable experimental programs. Due to the large interference of different environments on the data, this module can perform preprocessing such as normalization processing on the hyperspectral data and electrical impedance data. The data is analyzed by various algorithms such as principal component analysis and support vector machine. Finally, combined with the corn seed spectrum-vigor and electrical impedance-vigor models, the seed vigor information was obtained, graded and marked.

Preferably, the conveyor belt is provided with guide rails perpendicular to the conveying direction, and the detection board is slidably fitted on the guide rails. The detection board is controlled by a computer, moves laterally according to the vigor classification of the corn seeds, and controls the corn seeds to fall into the corresponding holding tanks.

The detection board can move in the direction perpendicular to the movement of the conveyor belt, and moves accordingly according to the received grading signal, so that corn seeds of different vitality levels fall into different storage tanks in the collection container at the end of the conveyor belt, realizing efficient online screening.

Description of drawings

Figure 1 is a device for online detection of corn seed vigor based on hyperspectral imaging and electrical impedance;

Figure 2 is a side view of the hyperspectral imager;

Fig. 3 is the electrode structure of a detection unit;

Fig. 4 is a flowchart of a method for detecting and screening corn seed vigor.

Detailed ways

The present invention will be described in detail below in conjunction with the embodiments and accompanying drawings, but the present invention is not limited thereto.

The device for detecting the vigor of corn seeds based on electrical impedance as shown in FIG. Wherein, the hyperspectral imaging chamber 3 includes a CCD camera 8, two sets of imaging spectrometers (380-1023nm and 874-1734nm) 9, imaging lenses 10 of XNP 1.4/17 (380-1023nm) and OLES15 (874-1734nm), Fiber- Lite DC950 line light source 11 and dark box. The spectral resolution of 380-1023nm is 2.8nm, and the spectral resolution of 874-1734nm is 5nm.

Corn seeds to be tested are stored in the material box 5, and the bottom of the material box 5 is connected with a chute 12 capable of outputting a single seed. pass.

The detection board 7 is slidably installed on the guide rail on the detection board 5 , the detection board 7 is provided with detection holes 13 for receiving the seeds output by the chute 12 , and each detection hole 13 is provided with a pair of electrodes. The structure of the electrode pair is shown in Figure 2. The electrode pair is two electrodes with the same surface area and interlaced with each other. Each electrode includes an arc-shaped plate 16 connected to the electrode line 15. The arc-shaped plates 16 of the two electrodes are arranged symmetrically. The two arc-shaped plates 16 are provided with grating bars 17 that extend opposite and are staggered at intervals. The electrode pair in this embodiment is made of graphite.

As shown in Figure 3, the impedance analyzer 1 is connected to the electrode pair in the detection hole 13 through the electrode line 15, and the excitation current is applied to the seeds through the electrodes, and the seed vitality is detected and classified according to the impedance value and phase angle between the electrode pairs.

The collection container 7 is located at the tail end of one end of the conveyor belt, and is provided with at least two holding tanks for accepting corn seeds of different vigor levels.

The detection board 5 is provided with a guide rail 14 perpendicular to the conveying direction, and the detection board 7 is slidably fitted on the guide rail 14 . The detection board 7 is controlled by the computer 1, moves laterally according to the vigor classification of the corn seeds, and controls the corn seeds to fall into the corresponding holding tanks in the collection container 6.

The batch of corn seeds to be tested is put into the feed box 4, and the seeds enter the conveyor belt through the chute of the feed box 4. Because the size of the chute can only accommodate a corn seed of normal size, and the gap between its end and the conveyor belt 5 is less, it can be ensured that each detection hole of the detection plate 7 on the conveyor belt 5 just accommodates a seed. The computer 1 then controls the impedance analyzer 2 to apply a variety of suitable excitation currents to the corn seeds, and simultaneously detects the impedance value and phase angle between the electrodes, and the computer 1 automatically saves and records the data. After the computer 1 analyzes the electrical impedance data, the corresponding corn seeds are initially graded and marked. The corn seeds enter the hyperspectral imaging room 3 at a set constant speed along with the conveyor belt 5. The Fiber-Lite DC950 line light source 11 provides a suitable light source. Two sets of imaging spectrometers 9 and CCD cameras 8 perform hyperspectral imaging on the corn seeds. Since the conveyor belt 5 has been moving at a constant speed, the imaging data is continuously collected and stored in the computer 1 . The computer 1 performs preprocessing, feature analysis, model comparison and other analysis on the imaging data in real time, and then combines the results of the electrical impedance data analysis to perform secondary classification and labeling of the corn seed vigor. The detection board 7 moves in the direction perpendicular to the movement of the conveyor belt according to the received marking signal. That is, the detection board 7 containing the high-viability seeds moves upward, and the low-viability seeds move downward. At the end of the conveyor belt 5, the seeds can drop into the collection container 6 containing multiple holding tanks, so as to realize the online detection and screening of corn seeds with different vigor.

As shown in Figure 4, the detection method of the present embodiment comprises the following steps:

1) Take the batch of corn seeds to be tested as a sample, use the material box to make the seeds evenly distributed on the conveyor belt 5, and each detection hole 13 accommodates one seed;

2) The electrodes in the detection hole 13 are connected to the impedance tester 2, and the computer 1 controls the impedance tester 2 to apply a variety of different suitable excitation currents to the corn seeds, and simultaneously detect the impedance value and phase angle between the electrodes. Computer 1 automatically records data and carries out real-time online analysis to the collected data, and carries out preliminary evaluation and classification to corresponding corn seed vigor;

3) The corn seeds enter the hyperspectral imaging chamber 3 at a constant speed along with the conveyor belt 5, and the hyperspectral imager performs real-time imaging of the corn seeds through the experimental program, and the imaging data is automatically stored in the computer 1 for online data analysis;

4) The computer 1 comprehensively analyzes the electrical impedance data and the hyperspectral imaging data, conducts secondary evaluation and grading of the vigor of corn seeds, and marks corn seeds of different vigor levels, and controls the detection plate 7 on the conveyor belt 5 to move on the vertical conveyor belt. Move accordingly in the direction of motion;

5) The seeds fall into different holding tanks at the end of the conveyor belt 5, and the detection board 7 is reset;

6) Repeat steps 1)-5) to realize on-line detection and screening of the seed vigor of corn seeds.

The above descriptions are only examples of the preferred implementation of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention within.

Claims (4)

1. The utility model provides a maize seed vigor detection device based on hyperspectral imaging and electrical impedance which characterized in that includes:

the corn seed box is used for storing corn seeds to be detected, and the bottom of the corn seed box is connected with one or more chutes for outputting single seeds;

the detection plate is provided with detection holes for receiving the seeds output by the sliding chute, and each detection hole is internally provided with an electrode pair; the electrode pair is two electrodes which have the same surface area and are mutually staggered; each electrode comprises arc plates connected with the electrode wires, the arc plates of the two electrodes are symmetrically arranged, and the two arc plates are provided with grid bars which extend oppositely and are staggered at intervals;

the detection plates are cuboid, and a plurality of detection plates are sequentially spliced on the conveyor belt; the conveying belt is provided with a guide rail vertical to the conveying direction, and the detection plate is in sliding fit with the guide rail;

an impedance analyzer connected to the electrode pairs for applying an excitation current to the seed through the electrodes according to electrical impedance data between the electrode pairs;

the hyperspectral imaging device is arranged above the conveyor belt and is used for imaging the corn seeds in the detection holes in real time;

the computer is used for comprehensively analyzing the electrical impedance data and the hyperspectral data and grading the vitality of the corn seeds;

the collecting container is positioned at one end of the conveying belt, and at least two accommodating grooves are formed in the collecting container and are used for receiving corn seeds with different vigor grades;

the detecting plate is controlled by a computer and transversely moves according to the vitality of the corn seeds in a grading way to control the corn seeds to fall into the corresponding accommodating grooves.

2. The hyperspectral imaging and electrical impedance-based corn seed vigor detection device of claim 1, wherein the electrode pair is made of graphite.

3. The apparatus according to claim 1, wherein the hyperspectral imaging and electrical impedance based corn seed vigor detection apparatus comprises two sets of imaging spectrometers with wave bands respectively ranging from 380 nm to 1023nm and from 874 nm to 1734 nm.

4. The hyperspectral imaging and electrical impedance based corn seed vigor detection device of claim 3, wherein the surfaces of the conveyor belt and the detection plate are black non-reflective materials.

Images