NL1002966C2 - Method of determining quality of seed of lot of crop - involves determining proportion of seedlings which have leaf surface area that exceeds surface area threshold, generating indication of seed quality based upon proportion determined - Google Patents

Abstract

The method involves generating an image of a number of seedlings grown from a number of seeds selected from a seed lot. The leaf surface area of each of the seedlings is determined from the image. The total leaf surface area of the seedlings is determined from the image. A surface area threshold relating to the leaf surface areas of a number of the seedlings is determined. The proportion of the seedlings which have a leaf surface area that exceeds the surface area threshold is determined. An indication of seed quality is generated based upon the proportion determined.

Description

Title: Method for producing a quality indication, which represents the quality of a batch of seed.

The invention relates to a method of producing a quality indication that indicates the quality of a batch of seed, comprising the following steps: (a) selecting a number of seeds from the batch of seed; and (b) growing a number of seedlings from the seeds selected in step (a) until the seedlings have leaves. A number of tests play a role in such methods.

10 Seed Growing Strength Tests

In 1876 Nobbe described in "Handbuch der Samenkunde" his observations regarding the differences in seed vigor and "germ energy". Since that time, the nature and importance of seed vigor has steadily increased.

15 Today, the Vigor Test Committee of the

Association of Official Seed Analysts accepted the following definition of seed vigor: "Seed vigor includes those traits that determine the ability to grow uniform quickly and develop normal seedlings over a wide range of field conditions." Seed Vigor Testing Handbook. published by the Seed Vigor Testing Committee of the Association of Official Seed Analysts (1983)). (AOSA).

Biologically, seed vigor is based on the genetic constitution of seeds that establishes their maximum physiological potential, based on the fact that seeds deteriorate in maturity and this deterioration continues until all seed tissues are dead. The degree of deterioration, including the loss of vigor, is determined not only by heredity but also by events occurring during seed development, harvesting, processing and storage.

Several categories of seed growth force tests are known. These categories are: (1) seedling growth and evaluation tests (often referred to as "seedling growth vigor classification and seedling growth rate" tests); (2) load tests; and (3) biochemical tests.

Testing germination.

In addition to vigor testing, germination testing is often performed to determine seed quality. Germination (in laboratory practice) is defined by the AOSA as "the growth and development from the seed embryo of those essential structures, which for the seed concerned indicate the ability to produce a normal plant under favorable conditions ". Germination test results establish the maximum plant yield potential of a batch of seed and correlate very well with growing under favorable field conditions.

25 Today, the germination test is the main accepted criterion for determining seed viability. The test results are typically obtained from seeds placed under favorable germination conditions. Essentially, germination tests are made on artificial, standardized, essentially sterile media, in moist, temperature-controlled germinators for periods long enough to germinate fairly "weak" seeds. However, a problem with this known 1002966-3 germination test is that this test overestimates field outgrowth because favorable conditions in the field are rarely, if ever, present.

Another problem with the known germination test is that this test scale is also dimensionless. Seed does not germinate or germinate. Therefore, every seed to be germinated is by definition equal in plant-producing capacity. The results of a germination test are expressed as a percentage from 1 - 100%. The problem with using such percentages is that this is very misleading. For example, a batch of 50% germinating seed should produce the same rebellion as a batch of seed that germinates 100% provided twice as much seed is planted. In some situations this could be the case, but usually this will not be the case.

15 Types of vigor test ♦ a. Seedling growth and evaluation tests:

Some vigor tests are performed under the same conditions as the standard germination test except that seedling growth is measured or evaluated in different ways. Seedling growth and evaluation tests are generally inexpensive and relatively fast. However, the drawbacks of these tests are that conditions are difficult to standardize between laboratories and the seed analyst must be able to determine if the seed has germinated.

The seedling vigor classification is equal to the standard germination test. The only difference between the two tests is that the normal seedlings are further classified as "strong" or "weak". A seedling is often characterized as weak if the primary root and / or cotyledon is missing, if it shows hypocotyl fractures, damage, necrosis, deformations or turns. Normal seedlings are characterized as "strong". Using • 0 0 2 9 6 6 - 4 - as this basis, seedlings are divided into seedlings with shortcomings and seedlings without shortcomings. Although this. test involves very little work, nonchalant • handling of the test may result in errors.

5 The seedling growth rate test involves a measurement of the seedling growth. In this test, seeds germinate according to the standard germination test with a more specific moisture content of paper towels. At the end of the germination period, seedling growth is measured. Linear growth and dry weight are usually determined. Seeds producing a single straight shoot or root can be measured to determine linear growth. The seedling growth rate test has four limitations: (1) measuring the seedling and removing the cotyledons or other storage tissues prior to oven drying are relatively time consuming; (2) the seedling rack can be very different in culture varieties; (3) the germination rate is influenced by moisture and temperature; and (4) the seed size affects hypocotyl growth in soybeans.

20 b. Load tests:

Different types of load tests are known. Some load tests simulate a load that seeds experience in the field. The theory behind the load test is that under sub-optimal or load conditions, seeds with a high vigor have a greater potential for outgrowth.

In the accelerated aging test, e.g. seeds placed at a temperature of 40 - 45 ° C and almost 100% relative humidity for different periods of time, after which a germination test is performed. This test is relatively inexpensive.

The cold test simulates early spring field conditions by establishing a high topsoil humidity and a low crop soil temperature. The seeds are placed in 1002966-5 topsoil in a plastic box or in paper towels with topsoil and grown at 10 ° C for a specific period of time. At the end of the cold period, a favorable temperature for germination is changed. The out-5 growth rate is considered an indication of the seed growth power. However, there is a problem with microorganisms in the cold test. Micro-organisms often cause a deterioration of the seed, fungi and other problems. In addition, specific crop conditions are difficult to standardize from laboratory to laboratory.

The cool germination test involves the germination of seeds in the dark at consistently low temperatures such as 18 ° C for several days. In essence, this test is a seed output 15 test. This test is also referred to as the slant plank test, which has been used to predict field vigor in lettuce, carrots, cauliflower seeds and cotton.

See O.E. Smith et al., "Studies on Lettuce Seed Quality: I. Effect of Seed Size and Weight on Vigor," J. Amer. Soc.

20 Hort. Sci. 98 (b): 529-533 (1973). McCormac, A.C. et al., "Automated Vigor Testing of Field Vegetables Using Image Analysis", Seed Sci. and Technol. 18: 103-112 (1990) -.

c. Biochemical tests:

Biochemical tests measure certain metabolic phenomena in seeds associated with germination and can be used to investigate vigor.

The tetrazole test measures the activity of the dehydrogenase enzyme. These enzymes reduce tetrazole chloride salt, which is colorless, to form a water-insoluble red compound, formazone, which gives living cells a red color. The dead cells remain colorless. See the Seed Vigor Testing Handbook, published by the Seed Test Committee of the Association of Official Seed Analysts 1002966 (1983).

- 6 -

Conductivity tests include measuring the conductivity in seepage water. Seeds of low vigor often have poor membrane structure and leak for a long time. Seeds with such a poor membrane structure often lose electrolytes, such as amino acids and organic acids, when they soak up water and thereby increase the conductivity of the trickle water.

Image analysis.

10 Image analysis, also known as machine vision, is a computer-based system used in the plant industry. The most common components of an image analysis system are a camera, a frame gripper to digitize the analog image and store it in a 15 RAM, a computer to run image processing, image analysis classification and user access software, and information output hardware such as a monitor and printer.

See Draper, S.R. et al., "Machine Vision for the Characterization and Identification of Cultivars", Plant Varieties 20 and Seeds 2: 53-62 (1989). Image analysis provides a new way of studying and analyzing plants and seeds. Image analysis is e.g. used to analyze and record the shape of plant organs and seeds Draper, S.R. et al., "Preliminary Observations with a Computer Based System for Analysis of the Shape of Seeds and Vegetative Structures", J. Nata. Inst. Agric. Bone. 36, 387-395 (1984); Travis, A.J. et al., "A Computer Based System for the Recognition of Seed Shape", Seed Sci. & Technol.

13: 813-820 (1985). Image analysis is also used to determine the shape and size of plants to classify, characterize, identify and record new plant varieties. See Keefe, P.D. et al., "An Automated Machine Vision System for the Morphometry of New ί V 0 1 9 6 6 - 7 -

Cultivars and Plant Gene Bank Accessions "; Draper, S.R. et al.," Machine Vision for the Characterization and Identification of Cultivars ", Plant Varieties and Seeds 2: 53-62 (1989).

5 Image analysis has been used to select visually healthy plants from a series of growing specimens and to transfer them to separate growth pots. After transferring to the individual growth pots, image analysis is used to further observe the plantlets until they develop into salable specimens. See He, W.B. et al., "Processing of Living Plant Images for Automatic Selection and Transfer", Computers and Elecrtronics in Agriculture, 6: 107-122 (1991).

Image analysis is also used in determining the leaf anatomy. Blade surfaces and blade lengths are e.g. measured by image analysis in leaves obtained from the wild ARABIDOPSIS THALIANA.

See Pyke et al., "Temporal and Spatial Development of the Cells of the Expanding First Leaf of ARABIDOPSIS THALIANA 20 (L.) Heynh", J. of Exp. Bone. 42: 1407-1416 (1991).

Image analysis has also been used in growth vigor and germination testing · Image analysis has been used to measure the results of the inclined plank test, the accelerated aging test and the cold test. See 25 Keys R.D. et al., "Automated Seedling Length Measurement for Germination / Vigor Estimation Using ACASAS (Computerized Automated Seed Analysis System)", J. of Seed Technol. 9: 40-53 (1984). McCormac, A.C. et al., "Cauliflower (BRASSICA OLERACEA L.) Seed Vigor: Imbibition Effects", 30 J. of Exp. Bone. 41: 893-899 (1990); McCormac, A.C. et al., "Automated Vigor Testing of Field Vegetables Using Image Analysis", Seed Sci. & Technol. 18: 103-112 (1990).

The problem with the known seed vigor tests is 1002966 - 8 - that most tests are done under specific, controlled growth conditions or use chemicals to investigate the vigor of the seedlings. Therefore, a special treatment of the seed has to take place to determine the vigor class of the seedlings.

In addition, most seed vigor and germination tests must be performed by licensed seed technologists. In order to become a recognized seed technologist, a number of specific requirements set by the Society of Commercial Seed Technologies must be met (these requirements are known to experts and will not be discussed in detail).

The program to meet these requirements is supported and promoted by the United States Department of Agriculture. To e.g. To certify seed for export, a licensed seed technologist must have performed the Germination Test.

The vigor classification according to the invention does not require the seeds to grow under specific controlled growth conditions. According to the invention, the vigor classification can be determined in seedlings growing under any type of growing conditions. E.g. the vigor classification of seedlings growing under optimal conditions, greenhouse conditions or outdoor conditions can be determined.

In addition, in the seedling growth rate test, each growth factor (root size, leaf size, stem size, etc.) is measured as part of the test. The idea is that the party that grows the most from day to day has the most vigor. In the invention, growth vigor classification is determined by testing the seedling at each stage of development as long as the leaves are present and do not overlap.

In addition, the known seed vigor and 1002966 9 germination tests require the presence of a licensed seed technologist to perform vigor testing.

A grower who cultivates seed trays must therefore have an accredited seed technologist to determine the 5 vigor classification of his seedlings.

According to the present invention, the presence of a recognized seed technologist is not necessary. Alternatively, anyone can be trained to perform growth vigor testing according to the invention.

The present invention therefore provides a means by which a grower can determine the vigor classification of a seedling filler tray grown under a variety of growing conditions and at any stage of development.

To this end, the method according to the invention is characterized by the following steps: (c) orienting the seedlings with respect to an image generator with a field of view such that surfaces of the leaves are located in the field of view of the image generator; (d) generating an image of the number of seedlings grown in step (b) and oriented in step (c), so that the image contains leaf surfaces; (e) automatically taking a number of measurements from the image of the seedlings generated in step (d), the measurements being measurements of leaf areas of the seedlings; and (f) assigning a quality indication to the batch of seeds from which the seeds were selected in step (a), the quality indication based on the measurements taken automatically in step (e).

The invention advantageously relates to a method for examining the quality of a sample batch of seed for a special harvest. The quality of a batch of seed is examined by determining the vigor classification for that particular crop of seedlings. Once the quality of the seed batch has been examined, this examination can be registered.

1002966 10

To determine the vigor classification, seeds from a particular crop must be sown in a suitable medium sufficient to support growth. The seeds are grown in a filling tray or outside in the field.

5 If the seeds are grown outdoors in the field, a photo of the seedlings is taken at an appropriate stage of development to determine their vigor classification.

The seeds can be grown under optimal conditions, commercial conditions, outdoors, or in a greenhouse.

Therefore, according to the present invention, the growth vigor classification determined is not dependent on the growth conditions.

Image analysis is used to determine the vigor classification. The image analysis-15 equipment is used to measure the total area of the leaves of the seedlings in a filler tray, Petri dish or photo, the standard deviation of the leaf area of the seedlings, and the germination percentage of the seedlings. When these measurements and calculations have been made, the vigor classification of the seedlings is determined by dividing the total area of the leaves by the standard deviation of the leaves multiplied by the germination rate.

The invention will now be explained in more detail below with reference to the drawings and the description.

Fig. 1 is a flowchart of a relevant part of the computer program used with the image analysis equipment, more particularly a control computer, to determine the vigor classification of a sample seedling hopper with 30 seedlings. The person performing the growth vigor test works with a control computer to determine the growth vigor classification.

Fig. IA is a flow chart of steps performed when placing seedlings under a video camera.

100 2966 11

Fig. 2 is a printout of a control monitor screen after determining the vigor classification for a seedling filler tray. The figure shows that a control computer measured the area of the leaves of the 5 seedlings, and calculated the standard deviation of the leaves and the germination percentage.

Fig. 3 is a print-out of a control monitor screen after determining the vigor classification for a seed box of petunia plants.

FIG. 4 is a print-out of a control monitor screen after determining the vigor classification for a filler tray with different impatience plants.

Fig. 5 is a print-out of a control monitor screen after determining the vigor classification for a filler tray with impatience plants.

Fig. 6 is a printout of a control monitor screen after determining the vigor classification for a pansy filler tray.

Fig. 7 is an example of a computer program that can be programmed into a control computer to practice the method of the invention.

Fig. 8 is a control monitor screen print showing the "color detection" window and graphs for color, saturation and intensity. The figure also shows 25 bars to the left of the color and saturation graphs that can be set if the computer does not detect seedlings.

The vigor classification can be used to examine and indicate the quality of a sample batch of seed for a particular crop of seedlings. In general, the greater the vigor classification, the higher the quality of the seed batch. The vigor classification is important for commercial growers from the inventory control point of view. Presumably, low-vigor classification seed trays and therefore low quality seed lots will not be sold or used.

The vigor classification can also be used to compare the quality of a particular batch of seed from a crop 5 grown in one year with the quality of a batch of seed from the same crop grown in another year or in another place or by a commercial seed breeder. Also, the vigor classification can be used to measure the quality of a batch of seed from a crop grown in one month (e.g. December) and the quality of a batch of seed from the same crop grown a few months later {e.g. April of the following year).

To determine the vigor classification of a sample-fill-15 seed tray with seedlings and finally the quality of a particular batch of seed from which seedlings have been taken, seeds are preferably sown in a fill seed tray. A fill seed tray or tray is a cultivation tray made from a polymer, plastic or wood. The filler tray is divided into sections of a certain size called "cells", which contain a medium capable of supporting plant growth. Filling trays have e.g. dimensions of approximately 28 x 56 cm and can contain 80-800 cells. Seeds are then sown in the medium. The medium used in the filler tray can be any type of medium such as topsoil, peat moss, water or gel. When water or gel is used, the medium must be supplemented with sufficient nutrients to support seedling growth. Preferably the medium is 30 peat moss - vermiculite - perlite.

Preferably, the seeds are sown and grown in a filler tray. However, seeds that are not grown in a filler tray can also be tested. E.g. the quality of batches of seed grown in a field or garden can also be determined. For seeds grown in such non-steered environments, a photo such as a slide or a digital image, such as a video tape, is taken from the seedlings at an appropriate development stage. The photograph or digital image is then used for vigor analysis testing purposes in place of the potting tray during testing. Seeds grown in a laboratory in a test tube or Petri dish can also be evaluated.

Any type of seed can be tested using the present invention. E.g. flower seeds, such as petunias, violets, or impiens can be used. Vegetable seeds such as cucumber, beans, lettuce or paprika can be used. Fruit seeds such as strawberries, tomatoes, pumpkins, apples and oranges can be used. Agricultural crop seeds, such as grain, wheat and barley can be used. Grass and turf seeds can be tested provided they are obtained at the correct stage of development. Grass and 20 turfs can be evaluated in digital format. The vertical orientation of the seedling from the topsoil to the top of the seedling can e.g. be used to obtain the image. Pre-sprouted or non-sprouted seeds may be used.

Once sown, the seeds are grown under any type of growing conditions. The seeds can e.g. be grown under optimal conditions or under commercial greenhouse conditions or can be grown outdoors in gardens or fields or in a greenhouse etc. However, regardless of where the seeds are grown, sufficient water, aeration, oxygenation, and lighting must be provided so that the seeds will germinate.

The conditions under which the seedlings are grown do not affect the vigor classification. Regardless of the conditions under which the seedlings are grown, the vigor rating can be used to determine which lots of seed have the most vigor. E.g. a batch of seed grown under optimal conditions can be tested followed by a seed tray grown under normal outdoor conditions. This variation in test conditions is unique in the field of vigor testing.

The known vigor tests test the determination of the vigor of plants grown under optimal or less than optimal growing conditions. According to these tests, the culture conditions were considered to be very significant. However, the vigor classification of the invention does not depend on the culture conditions.

When the seedlings appear and the leaves fold outwards, the vigor classification can be determined. It is preferable that the vigor classification can be determined before the leaves form a canopy or begin to overlap significantly so that accurate results can be obtained.

The vigor classification is determined using image analysis equipment. The image analysis equipment performs a series of measurements and calculations used to determine the vigor classification for the sample seeds and then for the associated batch of seed. The image analysis or machine vision equipment preferably includes a control computer, a control monitor, a video camera and software. Other equipment that facilitates testing such as color modules, image memory, expansion modules, etc. can be used.

Preferably, the control computer is any suitable PC-compatible control computer. Such control computers are known to those skilled in the art and will not be described in detail. The software program that can be used 1002966-15 is preferably Quantimet Image Processing Software (QUIPS). With QUIPS, a routine sequence of instructions can be generated to perform repetitive image analysis tasks. More specifically, QUIPS enables an image 5 analysis option that can record a sequence of image analysis operations (called "routine"), and which can run on a computer. The routine can be used to generate application solutions that can be repeatedly used by other users of the system 10 without any specialist knowledge of the operation of the system. In addition to the above equipment, a printer is usually required. Any suitable computer printer can be used.

When the seedlings are ready for testing, the filler tray or photo is placed under the video camera. A black and white or color video camera can be used. The video camera is used to produce a direct image of the sample seedlings.

The video camera sends the image to the control monitor.

20 The control monitor has an image of a stationary grille on its screen. The stationary grid is in the form of small boxes or cells. The direct image is in line with the stationary grid. More specifically, the image is aligned so that each seedling is within only one of the boxes or cells.

Sometimes the direct image will have some spots, speckles or other distortions. In those cases, the computer is used to smooth and / or modify the image so that a clearer, more easily analyzed image can be obtained. E.g. spots and speckles can be removed by looking at the pixels surrounding the spots and speckles and instructing the computer to give the spots and speckles the same color as the surrounding pixels or by the computer specifying the spots, speckles and surrounding spots. pixels mean to the same color.

The live image of the seedlings on the monitor is used to produce an image of the seedlings. From the generated images, the control computer calculates the growth force classification.

More specifically, from the generated images, the control computer measures the total area of all 10 leaves in the seed box or photo. The measurement can be performed in mm2. The computer then calculates the standard derivation of the open leaf area (in mm2) and the germination percentage of the seedlings. The computer calculates the germination percentage by calculating the total seedlings in the grid and dividing this by the number of squares or cells in the grid. This number is then multiplied by 100 to give the germination percentage.

The germination percentage can be calculated in a number of ways. E.g. the germination percentage can also be determined by having the computer look at each seedling separately in a cell. The computer then calculates the leaf area of each seedling and compares it to the leaf area of a commercially useful seedling from the same crop. The computer then determines the percentage of seedlings with a leaf area of at least 70% of the average area of the commercially useful seedling.

In addition, the computer can be programmed to take into account certain characteristics known to exist in a particular crop when the germination percentage is calculated. It is e.g. known to produce albino seeds. Albino seeds result in seedlings with white leaves. The computer can be programmed to count or not seedlings with white leaves -3 0 2 9 6 6 - 17 - when the germination rate is determined. A similar procedure can be used for plants that are too big or too small, too yellow, etc.

The vigor classification is determined by dividing the total area of the leaves by the standard deviation of the leaves and multiplying it by the germination percentage. The formula for calculating the growth vigor classification can be introduced into the programming of the computer or, instead, calculated by hand using a calculator.

A computer program that can be used for the method of the present invention can be manufactured by a computer programmer familiar with it. Fig. 7 is e.g. a computer program written with the 15 QUIPS software. This program can be programmed in a control computer and used in the method according to the invention.

If the program shown in Fig. 7 is programmed in a control computer, the method according to the invention is carried out as follows:

The person performing the testing works with the image analysis equipment to provide the information and parameters necessary to calculate the growth vigor classification. The first step is for the person performing the test 25 to place a filler tray, photo or digital image of the sample seedlings under a video camera. The camera does not need to have a specific distance from the hopper. It is only required that the camera and lens have the seed tray in the focus. When the seed tray is at the focal point 30, calibration is done by measuring with a ruler how much area is covered by a pixel. The pixel measurement is usually done in mm, although other units of measurement, such as cm, can be used. If e.g. The 1002966-18 program shown in Fig. 7 is used, covering a pixel 0.409982 mm (see Fig. 1A). The hopper can be placed on a photo tripod under the camera or placed on a guide system that moves below the camera. After the hopper has been placed under the video camera, the user performing the test will place himself in front of the control computer. Fig. 1 is a flow chart of the relevant part of the computer program used for interaction between the programmed control computer and the user. Preferably, the control computer is programmed with the above-described software, QUIPS.

The user accesses the system by using a password, key or the like used in conjunction with an access control system. After verifying that the user is authorized to access the image analysis system, the system displays a "Display" window and a "Tools" window. The user checks whether he is using the "QUIPS" software. The user then goes to the "Tools" window and chooses "Run". The system displays a "Rou-20 tine Header" window. The user then selects "File". Once the file window is opened, the user selects "Open". The user chooses the file for the specific crop and variety being tested, such as for petunias, impatiens, violets, etc. The computer then displays a "Routine Header" window with the selected file name at the top of the window. If the file name is incorrect, the user must start over and choose the correct file for the specific crop and variety being tested.

30 If the file name is correct, the user then selects "Continue". The system can display the "Feature Histo-gram" window. If the "Feature Histogram" window is displayed, the user checks the numbers of the X-axis and the Y-axis of the histogram to ensure that these numbers are correct for the harvest being tested. If the numbers are not correct, the user can enter the correct numbers. If the numbers are correct 5, the user then selects "Continue".

Then the computer gives a direct view of the seedlings in the seed hopper and a "Camera Set Up" window. Now the user can set various properties on the camera, such as aperture, zoom and focus, if necessary to improve the direct image. When the user is finished with the "Camera Set Up" window, the user presses "Continue".

After the "Camera Set Up" window, the computer displays a window containing a stationary grille calibrated in mm along with the direct image. The user aligns the seedlings in the hopper with the stationary grid. When the bin is properly aligned so that a full and clear image of the entire bin is properly positioned relative to the grid, the user presses "Continue".

The computer then displays an image of what has been detected as plant material from the obtained image. The computer also displays a “Color Detect” window that contains stored numbers for color, saturation, and intensity for the particular crop, and graphs for color, saturation, and intensity based on the direct image. The user checks the stored numbers for the color, saturation and intensity. If stored numbers are correct for the particular crop and the variety being tested, the user can then enter the correct numbers.

The user should now note that the computer has only partially detected some or all of the seedlings. If this is the case, the user can set the bars to the left of the color and saturation graphs (see fig. 8).

In addition to detection problems, the user may note that some processing of the detected image is required. If so, the user can use the "Binary Edit" window found in the "Tools" window under "Edit". The "Binary Edit" window has several "Editing" and "Undo Enditing" properties. For example, if the user notices that two seedlings in the image overlap or touch, he can select "Cut" from the "Binary Edit" window. The use of the "Cut" feature allows the user to draw a line between two detected seedlings and thereby instruct the computer to treat them as separate objects.

The user can also choose "Draw" in the "Binary Edit" window to draw a sketch of a seedling, which was not detected by the computer but is visible on the resulting image. The user can also omit an entire object that has been detected using the "Delete" function. The "Delete" function is also found in the "Binary Edit" window and is useful when perlite or vermiculite, i.e. culture medium particles, have been mistakenly detected by the computer as plant material.

The user can also use the "Erase" function in the "Binary Edit" window to draw a circle or rectangle around an object to be erased. The "Edit" function is useful if two seedlings are in the same cell and the user wishes to omit one from the obtained image.

The "Binary Edit" window also includes an "Undo" function that deletes all or part of an edit. If the "Undo" function is used, the user must again enter 1002966 - 21 - the entire operation after selecting "Undo".

When the "Editing" operation is completed, the user presses "Continue". Then the computer displays an "Image Utilities" window. The user then presses "Masked Copy" and then "Ok". Pressing "Masked Copy" produces a black and white image of the direct image.

After the "Image Utilities" window disappears, the computer displays an "Input Loop" window. The user can now use the "TV and Masked Copy Functions" located in the display window to tap the direct image and the black and white image to determine if any problems occurred on the "Masked Copy". If problems occur, the user can enter the number "1" in the "Input 15 Loop" window to obtain a new image or he can enter the number "2" to redetect the acquired image. If the user does not wish to re-acquire a new image or redetect the acquired image, he can then press "Continue".

20 After the "Input Loop" window, the computer displays an "Input Sampid $" window. The user can then enter the batch number located on the seed tray, if necessary.

Then the computer displays an "Input Sow" window 25. The user can now enter the sowing date and inspection date. The sowing date is the date the seed was planted. The inspection date is the date when the vigor testing was performed.

The computer then displays an "Input Variety $" - 30 window. The user can enter the specific variety name (or other suitable description) of the harvest being tested.

The computer then displays an "Input Rep $" window 1002966-22. The user enters the "Rep" number. The "Rep" nuimer is the number of the specific repeat of the test.

Finally, the computer displays the "Germination Results" window. In the "Germination Results" window, the 5 computer displays the germination percentage calculated for the seed tray as well as the vigor rating for the specific batch of seed sampled and tested. In addition, the screen can also display two "Feature Histogram" windows. One of the "Feature Histogram" windows contains a bar graph showing how many seedlings are of particular size. The other window shows the measurement of the area of the leaves and the standard deviation of the leaves. The window can also contain different sizes, such as the average leaf area, the maximum sheet area in a cell, the minimum sheet area in a cell, the median sheet area, etc. (see fig. 2).

It is important to note that various program subroutines can be added to improve the generated image. E.g. one can use the "Gaussian" command which is a form of image smoothing that uses a Gaussian core to give less blur of the image. The "Laplace 1" command can be used to detect edges in an image by performing a "Laplace" transformation using four neighboring pixels. The "Laplace 2" command uses eight neighboring pixels.

When printing is complete, the user presses "Continue" and asks the system if the user wishes to go back to the stationary display grid and set up another seedling filler tray to test vigor or replace the system instead. leave.

The typical time required for the control computer 1002966-23 to determine the vigor rating for a seedling seedling tray is about 1 minute. It appears that no special requirements or training are required for the user. The user can e.g. are trained to perform the growth-5 power classification in about half an hour.

The vigor rating obtained by the computer is then evaluated against a standard vigor rating scale accomplished through correlation studies with other seedlings from the same crop 10 sown in other commercial facilities. The quality of the seed batch is then determined, depending on where the vigor classification falls on the scale. The quality of the seed batch is then registered.

The vigor rating can also be evaluated against the vigor rating of seed trays determined earlier in the day, week, etc.

Depending on the vigor classification, determined by the method of the present invention, a particular batch of seed may or may not be sold or may be provided with an indication of a lesser vigor classification. For example, a vigor rating that does not give a grower 90% usable seedlings may be considered commercially unacceptable or may give rise to a lower price for the seeds when they are sold. This 90% percentage was developed based on information from other greenhouses. Vigor classifications are then correlated with separate commercial facilities. Since the conditions within a greenhouse are the same for all varieties or batches of seeds grown within the greenhouse concerned and because the vigor classification is independent of specific critical conditions, a grower can then make correlations within his greenhouse that are valid from harvest to harvest and year to year. This is extremely beneficial for 1002966 -24- growers for the following reasons: Suppose a grower wishes to sow 100,000 plants. Suppose the grower knows the vigor classification and the quality of the seed batch. Based on this information, the grower can: (1) 5 sow all 100,000 seeds based on his confidence in the quality of the batch; (2) sow the 100,000 seeds plus a buffer based on the quality of the seed batch and how many seeds he expects to emerge; or (3) sowing double the amount of seeds.

It has now also been found that seeds collected from seedlings with a high vigor classification, as determined by the method of the invention, have a higher shelf life. More specifically, different varieties of seeds with different growth power classifications were collected and stored under normal seed storage conditions for 5 months. Then, the seeds were planted in a filler tray and their vigor rating was obtained each month using the present invention. The seeds were also tested using the known standard fill test. As a result, the inventor found that the higher the seedling vigor classification, as determined by the present invention, the higher and better the shelf life of the seeds collected from the resulting plants. This op-. However, stroke life test result could not be correlated to, i.e. determined from, the known vigor test.

The main problem with the known tests is that these tests only estimate whether or not the seed batch will be good in 6 months or 12 months. Since there is no quantified measure of the seed vigor, these known tests are essentially a "pass" or "fail" type of an aging test.

Image analysis systems using more than one camera at more than one angle have also been considered for use in the invention. Such image analysis systems generate three-dimensional images. However, if such a system is used, the blade area is measured in 5mm3 instead of mm2.

The invention will be further described by the following examples.

Example 1

Fig. 3 shows the results of a vigor test on 10 a filler tray of petunia plants, called "Madness Burgundy". This test was done on January 20, 1995. First, petuia seeds were sown in a filler tray and grown until seedlings appeared and the leaves unfolded . The hopper was then placed under a video camera.

The growth vigor classification was determined using a control computer programmed according to Figure 7. The results are shown in Figure 3, indicating that two "Feature Histogram" windows are displayed. One of the "Feature Histogram" windows displays the results of the measurements of the total area of the leaves i.e. 3310.94 mm2. The standard deviation of the blade surface is also shown and is 10.64 mm2. The "Germination Results" window displays the germination percentage of the container as well as its vigor classification. The growth power rating is indicated under the heading "BALL VIGOR INDEX" ("BALL", Ball Vigor Index, and BVI (Ball Vigor Index) are brand names). The vigor classification is given as 296. "Total cells measured" is the total size of the sample. In this example, this amounts to 126 seeds per 30 test. The "Total germinated cells" is the number of seedlings in the grid.

Example 2

Fig. 4 shows the results of a vigor test 1002966-26 on a tray of impatience plants called "Super Elfin Blush" (brand name). Such impurity seeds were sown, grown and placed under a video camera, as in Example 1. The control computer was programmed according to Figure 7. The test was performed on January 31, 1995. The surface of the leaves is 9199.47 mm 2. The standard deviation is 20.90 mm2. The germination percentage is 99%. The vigor rating is 437.

Example 3 Fig. 5 shows the results of a vigor test of another seed box "Super Elfin Blush" (brand name) impatience plants. As in Example 2, impatience seeds were sown, grown and placed under a video camera. The control computer was programmed according to fig. 7. The test was also performed on January 31, 1995. The surface area of the leaves is 9377.98 mm 2. The standard deviation is 13.20 mm2. The germination percentage is 96%. The vigor rating is 682, which is higher than the vigor rating determined for the filler tray of Example 2. Customers, who were offered a choice between the seed lots of this example and Example 2, chose the seed lots of this example because of the higher vigor classification. The seed lots of this example perform better for the grower than the seed lots of Example 2 under any type of culture conditions. The seed lots from this example can be stored longer and will germinate better than the seed lots from Example 2.

Example 4

Fig. 6 shows the results of the vigor test on a violet seed tray called "Maxim Marina" (brand name). Pansy seeds were sown, grown and placed under a video camera as in Example 1. The control computer was programmed according to Figure 7. The test was performed on 100 2 QS 8 - 27 - 2 December 1994. The surface of the leaves is 8781.45 mm 2. The standard deviation is 27.37 mm2. The germination percentage is 92%. The vigor rating is 295.

Although the invention has been described mainly in connection with special and preferred embodiments, it will be apparent to those skilled in the art that modifications are possible without exceeding the scope of the invention.

1002966

Claims (9)

A method of producing a quality indication indicative of the quality of a batch of seed, comprising the following steps: (a) selecting a plurality of seeds from the batch of seed; and (b) growing a number of seedlings from the seeds selected in step (a) until the seedlings have leaves, characterized in that the following steps are performed: (c) orienting the seedlings relative to of an image generator with a field of view such that surfaces of the leaves are in the field of view of the image generator; (D) generating an image of the number of seedlings grown in step (b) and oriented in step (c) so that the image contains surfaces of the leaves; (e) automatically performing a number of measurements from the image of the seedlings generated in step (d), the measurements being measurements of leaf surfaces of the seedlings; and (f) assigning a quality indication to the batch of seed from which the seeds were selected in step (a), the quality indication based on the measurements taken automatically in step (e). A method according to claim 1, characterized in that the seedlings grow under any type of growing conditions and produce a quality indication regardless of the particular growing conditions under which the seedlings grow. Method according to claim 1, characterized in that step (d) comprises producing an image of the seedlings in which the seedlings are arranged in a two-dimensional arrangement. 1002966 Method according to claim 1, characterized in that step (e) comprises performing a number of sheet area measurements. Method according to claim 1, characterized in that step (e) comprises performing a number of measurements, each of which is indicative of the leaf area of one of the seedlings. 6. A method according to claim 1, characterized in that step (d) comprises using an image generator to generate the image of the number of seedlings grown from the seeds. Method according to claim 1, characterized in that step (e) comprises using a control computer to automatically perform the number of measurements from the image of the 15 seedlings. A method according to claim 1, characterized in that step (d) comprises using a control computer to assign the quality indication to the batch of seeds from which the seeds were selected in step (a). Method according to claim 1, characterized in that step (d) comprises using a control computer to generate a vortex qualification for the seed batch based on the measurements, the vigor qualification not solely dependent on the percentage of seeds selected from the batch of germinated seeds. 1002966