JP5716234B2 - Harvest quality prediction system and harvest quality prediction method using cross-sectional images of rice to be harvested - Google Patents

Description

  The present invention relates to a harvest quality prediction system and a harvest quality prediction method using a cross-sectional captured image of a planned harvest rice that can be easily and accurately executed to predict the harvest quality of the planned harvest rice composed of brown rice or brown rice. is there.

  In the Agricultural Products Mutual Aid System, mutual aid money can be received for the decline in yield and quality due to disasters, but damage is declared by the farmers before harvesting, and the investigators such as mutual aid associations are raised in the field (field It is necessary to evaluate the degree of damage by conducting a survey.

  It is easy to understand that damage was caused by pest damage and lodging, so there is almost no omission of damage by farmers, but damage from the appearance of napped rice such as lack of sunshine or frequent occurrence of milky grains due to high temperature or dry wind If it is difficult to predict the situation, damages will not be declared and mutual aid money will not be received.

Considering that field surveys for damage assessment are conducted in a wide area after the damage report and before harvesting, it is necessary to report the damage at around 10 days before harvest.
For this reason, some ears of the field are collected at this time and the quality of brown rice is examined by an objective or grain discriminator, but the appearance survey of immature brown rice predicts the quality at maturity This is a difficult situation.

  In addition, there is an example in which, when collecting some ears of the field and investigating the quality of brown rice objectively, it is left to the expert appraisers, but in this case, the appraisal results of individual appraisers vary. It tends to occur, lacks objectivity, and further increases the cost of training appraisers.

  In the past, for example, a method has been proposed to predict the occurrence rate of milky white grains at harvest from the appearance quality assessment of unpolished rice before harvesting (refer to “Gunma Agricultural Technology Center”). However, with this method, the quality of the mature stage can be predicted to some extent when the accumulated temperature after heading is 750 ° C or higher (corresponding to about 7 days before early rice harvesting), but it is difficult to predict milky white grains. It is said that.

  Furthermore, the method of estimating the milk white grain incidence from the dry matter production of rice, the amount of accumulated carbohydrates, and the accumulated temperature after heading (for example, “Study on modeling and prediction of rice immature grain development”, Journal of the Crop Science Society of Japan 77 Volume, Annex 1, P148-149, 2008), but the necessary data collection is complicated and difficult to use in the agricultural field.

  Of the conventional prediction methods, the modeling method requires a lot of growth data such as dry weight and carbohydrates and weather data as described above, and the prediction accuracy is low when using brown rice before harvesting. However, it has not been used to predict the occurrence of milky white grains in individual fields.

Conventionally, a harvest-time quality prediction system using a cross-sectional image image of a planned harvest rice that can accurately perform the harvest-time quality prediction of the planned harvest rice made of glutinous rice or brown rice according to the present invention, There is no quality prediction method.
If the prior literature of the present invention is enumerated, it is not directly related to the present invention, but it is possible to cite Patent Document 1 as an apparatus related to component analysis rather than appearance quality such as brown rice and rice bran.
Patent Document 1 includes a sample storage chamber in which grain samples are stored, and a floodlight for the purpose of measuring the components of brown rice, white rice, and even brown rice, without damaging or degrading them. And a grain quality judging device comprising a measurement chamber having a light receiving port, and a casing in which the light projecting means and the light receiving means are disposed, wherein the light projecting means emits irradiation light including near infrared light. A light source, a reflector that reflects the emitted light emitted from the light source toward the measurement hole, one opening facing the light source and the reflector, and the other opening facing the measurement hole, and the light source and the measurement hole There has been proposed a grain quality evaluation apparatus comprising a reflecting cylinder disposed between the two.

  However, as described above, the device of Patent Document 1 is a device related to component analysis rather than appearance quality, and in the case of a grain quality evaluation device, it mainly uses component analysis of brown rice, rice bran, etc. Of course, it is not the harvest quality prediction system and harvest quality prediction method for the planned harvest rice.

JP 2008-175760 A

  The problem to be solved by the present invention is that it is not necessary to go through the eyes of the appraiser, the cost for training the appraiser does not increase, and the worker who performs the prediction work may have special technical knowledge. Harvest quality prediction system using harvested rice cross-sectional images that can accurately predict the harvest quality of the planned harvest rice, which is made of glutinous rice or brown rice, without any need. There is no prediction method.

The harvest-time quality prediction system using the cross-sectional image of the planned harvested rice according to the present invention is a large number of grains that are collected from the field on the collection day before the actual harvest date and stored in each hole of the grain support plate. A grain cutter that cuts the body of the sample grain of the planned grain to be harvested, and a planned harvest that arranges each cut surface of the sample grains of the multiple harvested rice that has been cut by the grain cutter in an exposed state Acquired by the rice grain arranging means, the imaging means for scanning each cut surface of the sample grains of a large number of planned harvested rice grains arranged by the planned harvest rice grain arranging means, and obtaining a cut surface image of each grain, and the imaging means Image analysis processing means for analyzing the cloudiness portion of the endosperm portion in the cut surface image of each grain with reference to classification category information designated in advance, and classifying into predicted sizing, predicted white immature grain, and undecided grain, and said prediction Sized, predicted white immature, undecided Analysis of has been clouded portions of the endosperm portion of the grain of the cut surface image, and the prediction sizing, prediction white immature grains, based on the number information of the undetermined grain, the predicted white immature grains occurrence rate is almost no change The predicted white immature grain generation rate for the collection date in a certain period close to the actual harvest date is calculated using the number of grains corresponding to the number of holes in the grain support plate as the denominator. A computer unit having a calculation means for calculating by a calculation as a numerator, and predicted white immature grains on a collection date in a certain period close to an actual harvest date in which the occurrence ratio of the predicted white immature grains hardly changes. The most important feature is that the generation ratio of white immature grains on the harvest date is predicted based on the calculated generation ratio .

According to the first aspect of the present invention, by using the grain cutter, the harvested rice grain arranging means, the imaging means, the computer device, and the output means having the above-described configuration, the occurrence rate according to the quality of brown rice before harvesting up to now only by simple operation. The ambiguity of the prediction that was caused by the judgment of the appearance, the cost and time that were caused by the appraiser, and the individual differences in the analysis results without incurring costs for the appraiser In addition, it is possible to improve the measurement accuracy of the predicted grain size, predicted white immature grain, and undecided grain of the rice to be harvested, and in particular to improve the calculation accuracy of the occurrence ratio of predicted white immature grains. It is possible to realize and provide a harvest quality prediction system using cross-sectional images of planned rice.

According to the invention described in claim 2, the incidence rate according to the quality of brown rice before harvesting up to now by simple operations using the grain cutter, the harvested rice grain arranging means, the imaging means, the computer device and the output means having the above-described configuration. The ambiguity of the prediction that was caused by the judgment of the appearance, the cost and time that were caused by the appraiser, and the individual differences in the analysis results without incurring costs for the appraiser In addition, it improves the measurement accuracy of predicted grain size, predicted white immature grains, and undecided grains for 8 classification categories of rice grains to be harvested . can be harvested during achieve quality prediction system for providing with a possible harvesting plans rice sectional captured image be improved.

According to the invention described in claim 3, by using the grain cutter, the harvested rice grain arranging means, the image scanner, the computer device, and the printer or the display unit having the above-described configuration, it is possible to classify the brown rice quality before harvesting by simple operations. The ambiguity of the prediction that was caused by the appearance rate being judged by the appearance, the cost and time that was incurred by the appraiser, and the individual differences in the analysis results without incurring costs for the appraiser reduce, and, predictions about eight classification categories specified in advance of the grain harvest expected rice sized, prediction white immature grains, improve the measurement accuracy for undetermined grain, especially the calculation of the occurrence rate of prediction white immature grains It is possible to realize and provide a quality prediction system at harvest using a cross-sectional image of a planned harvest rice that can also improve accuracy .

According to the invention described in claim 4, by using the grain cutter, the harvested rice grain arrangement means, the image scanner, the computer device and the printer or the display unit having the above-described configuration, The ambiguity of the prediction that was caused by the appearance of brown rice quality before harvesting was judged by appearance, the cost and time that were caused by the appraiser, and without the expense for the appraiser training , To reduce individual differences in analysis results, and to improve the measurement accuracy of eight types of pre-designated classification categories of grains to be harvested, predicted white immature grains, undecided grains, especially predicted white immature grains It is possible to realize and provide a quality prediction method at harvest time using a cross-sectional image of the planned harvest rice that can improve the calculation accuracy of the occurrence rate of rice.

FIG. 1 is a schematic perspective view showing a grain cutter and a planned harvesting grain arrangement means in a quality prediction system for harvesting using a cross-sectional captured image of planned harvesting rice according to an embodiment of the present invention. FIG. 2 is a schematic perspective view showing an imaging unit, a computer device, and an output unit in a harvest quality prediction system using a cross-sectional captured image of planned harvest rice according to the present embodiment. FIG. 3 is a schematic block diagram of a harvest time quality prediction system using a cross-sectional image of the planned harvest rice according to the present embodiment. FIG. 4 is an explanatory diagram showing grains in which the cut surface to be predicted grain size is entirely transparent in the present embodiment. FIG. 5 is an explanatory diagram showing a grain in which the vicinity of the center of the cut surface, which is a predicted white immature grain, is white and circular in this example. FIG. 6 is an explanatory diagram showing grains in which the vicinity of the center of the cut surface to be predicted white immature grains is turbid in the form of nuclei and the peripheral portion thereof is already transparent in this embodiment. FIG. 7 is an explanatory diagram showing a grain having a band-like cloudy portion at the center part of the cut surface as a predicted white immature grain in this example. FIG. 8 is an explanatory diagram showing grains that are clouded from the back side to the center, in addition to the outer periphery of the cut surface that is predicted white immature grains. FIG. 9 is an explanatory diagram showing grains in which only the vicinity of the back side or the ventral side of the outer periphery is clouded as predicted white immature grains in this example. FIG. 10 is an explanatory view showing a grain in which the cut surface as an undetermined grain is milky paste or milky white and opaque in this embodiment, and the seed coat near the outer periphery is transparent at the initial stage of ripening. It is. FIG. 11 is an explanatory view showing a grain in which the center part of the cut surface to be determined as the undetermined grain is transparent and the outer peripheral part is clouded in this embodiment. FIG. 12 is an explanatory diagram illustrating a cutting process of a grain to be harvested and a setting process on an imaging unit in a quality prediction system at harvest time using a cross-sectional image of the planned harvest rice according to the present embodiment. FIG. 13 shows a harvest quality prediction system (branch line for brown rice cross section observation) using a cross-sectional image of a rice to be harvested when the degree of occurrence of white immature grains is low, and a grain discriminator (not shown) It is a graph showing the occurrence rate of predicted white immature grains on a collection period in a certain period close to the actual harvest date in which the occurrence ratio of predicted white immature grains hardly changes, which is almost the same as the actual harvest date . , Showing the relationship between the percentage of brown rice that has been judged to be milky white and white dead rice by the grain discriminator and brown rice that has been judged to be predicted white immature grain and the number of days before harvesting (control zone; Koshihikari in Kagoshima Prefecture It is cultivated in the development center field, and the upper and lower lines of each symbol in the figure are standard errors). FIG. 14 shows the quality prediction at the time of harvest using a cross-sectional image of the planned harvest rice when the degree of occurrence of white immature grains is high (covering black cold chilled rice with a shading rate of 50% from 2 to 22 days after heading). The harvest date obtained using the system and a grain discriminator (not shown), i.e., the predicted white immature grain of the harvest date in a certain period close to the actual harvest date when the occurrence rate of the predicted white immature grain hardly changes . It is a line graph showing the rate of occurrence (meaning of each line is the same as in FIG. 13), and the rate of brown rice determined to be milky white and white dead rice by the grain discriminator and the predicted white immature grain It shows the relationship with the number of days before harvesting (shaded area; covered with black cold chilled rice with a shading rate of 50% from 2 to 22 days after heading).

The present invention does not require the eyes of an appraiser, does not increase the cost for training the appraiser, and does not require that the worker who performs the prediction work has special technical knowledge. The purpose of providing a quality prediction system at harvest time using cross-sectional images of the planned harvest rice that can accurately predict the harvest quality of the planned harvest rice, which is made of glutinous rice or brown rice, is the collection date prior to the actual harvest date. A grain cutter that cuts the body of a large number of sample grains of brown rice or brown rice that are collected from the field and stored in each hole of the grain support plate, and cut by the grain cutter A plurality of planned grains to be harvested and arranged in a state where each cut surface of the sample grains to be harvested is exposed, and each of the sample grains of a plurality of planned grains to be harvested arranged by the planned grain to be harvested rice Run through the cutting plane Imaging means for obtaining a cutting surface image of each particle and, and analyzed with reference to the eight classification category information in advance specified clouding portion of the endosperm part of each grain of the cut surface images acquired by the imaging means , Predicted sizing, predicted white immature grain, image analysis processing means for classifying into undetermined grains, and the cloudiness portion of the endosperm portion in the cut surface image of each grain that has been predicted sized, predicted white immature grains, and undetermined grains Based on the analysis results and the number information of predicted sizing, predicted white immature grains, and undecided grains , the prediction of the collection date in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes The calculation means for calculating the generation ratio of white immature grains by calculation using the number of grains corresponding to the number of each hole of the grain support plate as a denominator and the number of predicted white immature grains as a numerator,
Storage means for storing the processing results of the imaging means, image analysis processing means, calculation means, and output information creation means for creating graph information or table information relating to the occurrence rate of predicted white immature grains based on the calculation results of the calculation means And an output means for outputting the processing results of the image analysis processing means and the output information creation means as an image, a graph or a table, and the occurrence ratio of the predicted white immature grains hardly changes. A cross-sectional image of the planned harvest rice that is configured to predict the occurrence rate of white immature grains on the harvest date based on the calculated occurrence rate of white immature grains on the harvest date in a certain period close to the actual harvest date. In the harvest quality prediction system used, the eight types of classification category information specified in advance are pre-ripe rice grains that are to be harvested and are pre-ripe grains that have a transparent cut surface. Grain that is sized and is white and circular in the vicinity of the center of the cut surface is the predicted white immature grain that becomes milky white, and that the center of the cut surface is clouded in the form of nuclei and the periphery is already transparent Is a predicted white immature grain that becomes milky white or heart white grain, and a grain that has a band-like cloudy part at the center of the cut surface is regarded as a predicted white immature grain that becomes a heart white grain from the back side in addition to the outer periphery of the cut surface If the grain that is cloudy toward the center is the predicted white immature grain that becomes milky white or heart white, and only the backside or ventral side of the cut surface is cloudy, the backwhite or belly white Predicted white immature grains, and the cut surface is milky pasty or the whole milky white and opaque grains are undetermined grains, the center of the cut surface becomes transparent with starch, and the outer periphery becomes cloudy This is realized by the configuration in which the existing grains are undetermined grains.

  Hereinafter, the harvest quality prediction system and the harvest quality prediction method using the cross-sectional image of the planned harvest rice according to the embodiment of the present invention will be described in detail.

As shown in FIG. 1 and FIG. 2, the harvest quality prediction system using the cross-sectional captured image of the planned harvest rice according to the present embodiment is brown rice or glutinous rice harvested from the field on the harvest date prior to the actual harvest date. The grain cutter 1 which cuts the trunk | drum of the sample grain M of many harvest planned rice which consists of, and shows each cut surface of the sample grain M of many harvest planned rice cut | disconnected by the said grain cutter 1 The harvested rice grain arranging means 41 arranged in the state of being arranged, and the cutting planes Ma of the sample grains M of the many harvested rice grains arranged by the scheduled harvesting rice grain arranging means 41 are scanned to cut the cut faces of each sample grain M The image pickup means 51 such as a color image scanner for acquiring an image, the classification based on the classification category information specified in advance of each sample grain M acquired by the image pickup means 51, and the generation ratio of the predicted white immature grains hardly change. In fact Calculating the occurrence percentage of predicted white immature grains harvesting date in a period of time close to the harvest date, the computer apparatus 61 of making such a graph information or schedule information on the occurrence ratio of the predicted white immature grains, cleavage of the sample particle M A printer 81 as output means for outputting a plane image, a graph, a table or the like and a display unit 64 are provided.

  As shown in FIG. 1, the grain cutter 1 has a rectangular box shape with an open top, and a rectangular flat part 3 for placing a grain support plate 21, which will be described in detail later, is formed at the bottom. At the same time, the apparatus main body 2 in which the side wall portion 4 is erected from each side of the flat portion 3 and a pair of the main body 2 laid in parallel from the side wall portion 4 on one end side to the side wall portion 4 on the other end side. The guide bar 5, the linear rack 6 provided in the same direction as the guide bar 5 on the outer side wall 4 of the one guide bar 5, and the pair of guide bars at a position near the side wall 4 on the one end side 5, a movable body 7 having both side portions slidable from one end side to the other end side, a pinion shaft 8 mounted in a vertical arrangement at a position on the rack 6 side of the movable body 7, and the pinion shaft 8 and fitted to the rack 6 (not shown) A safety handle 14 made of, for example, a transparent material, which is detachably attached to the upper surface of the movable body 7 using screws 11, and the movement. A cutting blade holder 10 attached to the lower surface of the body 7, and a cutting blade 12 made of, for example, a commercially available cutter blade detachably attached to the lower surface side of the cutting blade holder 10 using screws 13.

  The cutting blade 12 supports the grain in conjunction with the moving body 7 that moves in the direction of arrow a in FIG. 1 by meshing with the rack 6 and the pinion while being guided by the guide bar 5 according to the rotation operation of the operation handle 9. It is configured to move in the direction of arrow a along the upper surface of the plate 21.

  The harvested rice grain arranging means 41 includes a grain support plate 21, an alignment aid 23 that covers the grain support plate 21 before accommodating the sample grains M and assists the alignment of the sample grains M, and after the cutting process. A transparent support body (transparent petri dish) 31 for placing the cut surface Ma of each sample grain M toward the image pickup means 51 on the grain support plate 21 in a state where each sample grain M is accommodated. Have.

  As shown in FIG. 1, the grain support plate 21 is formed in a substantially rectangular parallelepiped shape, and the upper surface portion 21a has, for example, 10 rows in the vertical direction and 10 rows in the horizontal direction in the illustrated embodiment. In the arrangement, for example, a total of 100 holes (concave holes) 22 for sample grains M are formed, and as shown in FIG. It is fixedly arranged on the flat part 3 in a state of being accommodated in a protruding state.

  Examples of the fixing of the grain support plate 21 on the flat portion 3 include an example in which a concave portion having a shape corresponding to the bottom shape of the grain support plate 21 is provided on the upper surface of the flat portion 3. it can.

The alignment aid 23 has a hole diameter corresponding to each hole 22 of the grain support plate 21 in a rectangular flat part 23a having a dimension slightly larger than the dimension of the upper surface part 21a of the grain support plate 21. In addition, a corresponding number of alignment auxiliary holes (through holes) 24 (for example, a total of 100 in the arrangement of 10 rows in the vertical direction and 10 rows in the horizontal direction) are provided, and the four sides of the flat portion 23a are vertically moved from the flat portion 23a. It is comprised in the shape enclosed by the side wall part 25 which protrudes.
And let the lower part rather than the position of the flat part 23a in the said side wall part 25 function as an object for fitting to the outer periphery of the said grain support plate 21, and it is above the position of the flat part 23a in the said side wall part 25. The space surrounded by the portion is configured to function as an alignment auxiliary space for the sample grains M.

  The transparent body 31 has a rectangular shape having a size slightly larger than the size of the upper surface portion 21a of the grain support plate 21, a flat portion 31a made of a transparent material such as a transparent synthetic resin, and four sides of the flat portion 31a. And the flat portion 31a can be covered in close contact with the upper surface portion 21a of the grain support plate 21.

  As shown in FIG. 3, the computer device 61 includes a program memory 62 storing a program for controlling the overall operation, a control unit 63 for controlling the entire system based on the program, and an output means including a liquid crystal display. The display unit 64 includes a printer interface 65 that outputs image data and the like to the printer 81, an interface 66 for the imaging unit 51, and a keyboard 67.

Further, the computer device 61 preliminarily designates the cloudy part of the endosperm portion in the cut surface image of each sample grain M acquired by the classification category information storage means 68 storing the predesignated classification category information and the imaging means 51. The image analysis processing means 69 for classifying the classified category information into the predicted sized particles, the predicted white immature particles, and the undetermined particles, and the predicted sized particles, the predicted white immature particles, and the undetermined particles. Based on the analysis result of the cloudiness part of the endosperm part in the cut surface image of each sample grain and the number information of the predicted sizing, predicted white immature grain, and undecided grain, the occurrence ratio of the predicted white immature grain is hardly changed the calculating means 70 for calculating the occurrence percentage of predicted white immature grains harvesting date in a period of time close to the harvest date, the imaging means 51, the image analysis processing section 69, the processing result of the calculation means 70 A storage means 71 such as a hard disk that 憶 has an output information creating means 72 for creating a graph information or schedule information on the occurrence ratio of the predicted white immature grains calculation result based on the calculation unit 70.

  Next, with reference to FIGS. 4 to 12, the pre-designated classification category information stored in advance in the classification category information storage means 68 will be described in detail.

  In the classification category information, the image information of the cut surface Ma of each sample grain M is classified into, for example, eight classifications.

  That is, in the classification category information storage means 68, a grain to be harvested that has not yet been matured as shown in FIG. A grain that is white and circular in the vicinity of the center of the surface Ma is set as a predicted white immature grain that becomes a milky white grain, and the vicinity of the center of the cut surface Ma as shown in FIG. The predicted white immature grains that become milky white grains or heart white grains, and grains that have a band-like cloudy portion at the center of the cut surface Ma as shown in FIG. In addition to the outer periphery of the cut surface Ma as shown in FIG. 8, the grains that are clouded from the back side to the center are assumed to be milky white grains or predicted white immature grains that become heart white grains, and the dorsal side or belly as shown in FIG. If only the outer periphery of the side is cloudy, it becomes a white background or a white background, respectively. Predicted white immature grains, the cut surface Ma as shown in FIG. 10 is a milk-like paste-like or entirely milky white and opaque, and grains in which the seed coat near the outer periphery is transparent at the initial stage of ripening are undecided grains, As shown in FIG. 11, classification category information is stored in which the center portion of the cut surface Ma is transparent and the outer peripheral portion is clouded is determined as an undetermined particle.

  Next, harvest of the planned harvest rice consisting of brown rice or glutinous rice collected from the field on the harvest date prior to the actual harvest date by the quality prediction system at harvest using the cross-sectional image of the planned harvest rice according to the present embodiment A series of flows in the time quality prediction process will be described.

  First, as shown in FIG. 12, the alignment aid 23 is placed on the grain support plate 21, and the flat portion 23 a is brought into close contact with the upper surface portion 21 a of the grain support plate 21. The auxiliary holes 24 are aligned with the holes 22 of the grain support plate 21.

Next, for example, 100 or more sample grains M of the planned harvest rice are introduced from above the alignment aid 23, and the alignment aid 23 and the grain support plate 21 are shaken in small increments.

  As a result, the individual sample grains M are in a state of being vertically immersed in the respective holes 22 of the grain support plate 21 through the respective alignment assist holes 24 of the alignment assisting tool 23, and extra sample grains M Remains on the flat portion 23a.

  Next, the alignment aid 23 is removed from the grain support plate 21 together with the remaining sample grains M. Thereby, for example, the grain support plate 21 can be obtained in a state where the sample grains M are accommodated in a state where the upper half protrudes upward from a total of 100 holes 22.

  Next, the grain support plate 21 in which 100 sample grains M are accommodated in each hole 22 is set on the upper surface of the flat portion 3 of the grain cutter 1.

Next, by rotating the operation handle 9 and moving the moving body 7 in the direction of arrow a in FIG. 1 while being guided by the guide bar 5, the cutting blade 12 is also placed on the upper surface of the grain support plate 21. Along the arrow a.

  As a result, the cutting blade 12 cuts from the grain support plate 21 the body of each sample grain M whose upper half protrudes upward, and the cut surface Ma of each sample grain M is exposed. State

  Next, the transparent body 31 is placed on the grain support plate 21 in which the cut surface Ma of each grain sample is exposed on the upper surface part 21a, and the flat part 31a is grain supported. After being brought into close contact with the upper surface portion 21 a of the plate 21, the grain support plate 21 and the transparent body 31 are turned upside down while being pressed down, and the flat portion 31 a of the transparent body 31 is placed on the lower side.

  Next, with the flat portion 31a of the transparent body 31 being placed on the lower side, these are placed on the imaging means 51, and the cut surface Ma of each sample grain M is placed on the transparent flat portion 31a of the transparent body 31. Through the imaging surface 52.

  Then, the imaging means 51 images the cut surface Ma of each sample grain M, and sends the imaging data to the computer device 61.

  The image analysis processing means 69 of the computer device 61 stores, in the classification category information storage means 68, the cloudy part of the endosperm portion in the cut surface image of each sample grain M acquired by the imaging means 51 by the program. Each of the sample grains M is analyzed by referring to the eight kinds of classification category information as described above, and classified into one of predicted grain size, predicted white immature grain, and undecided grain.

  The cut surface images of 100 sample grains M acquired by the imaging unit 51 and the analysis result of the image analysis processing unit 69 are stored in the storage unit 71.

Further, the calculation means 70 of the computer device 61 includes the analysis result of the cloudy part of the endosperm portion of the cut surface image of the predicted sized particle, the predicted white immature particle, the undecided sample particle, and the predicted sized particle, Based on the information on the number of predicted white immature grains and undetermined grains, the occurrence ratio of predicted white immature grains on the collection date in a certain period close to the actual harvest date when the occurrence ratio of the predicted white immature grains hardly changes The number of grains corresponding to the number of holes in the support plate is used as a denominator, and the calculation is performed by calculating the number of predicted white immature grains as a numerator. Information on the generation ratio of predicted white immature grains is also stored in the storage means 71.

  Further, the output information creation means 72 creates graph information or table information related to the generation ratio of predicted white immature grains based on the calculation result of the calculation means 70 so as to have the form shown in FIG. 14, for example.

  The graph information or the table information relating to the generation ratio of predicted white immature grains created by the output information creating means 72 is visually output by the printer 81 or the display unit 64, respectively. Further, the cut surface image of each sample grain M is also visually output by the printer 81 or the display unit 64 as necessary.

Here, with reference to FIG. 13 and FIG. 14, the occurrence of predicted white immature grains on the collection date in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains calculated by the calculation means 70 hardly changes. A specific example of the ratio will be described.

  FIG. 13 shows a harvest date (harvest date) obtained by using a harvesting quality prediction system using a cross-sectional image of the planned harvested rice according to the present embodiment and a grain discriminator (not shown) manufactured by the applicant of the present application. The line graph which shows the generation | occurrence | production ratio of the predicted white immature grain in the period from the 14th day before) to the actual harvest date is shown.

In FIG. 13, the line graph indicated by ● indicates the generation ratio of brown rice determined to be predicted white immature grains and the number of days before harvesting by the quality prediction system at harvest using the cross-sectional image of the planned harvested rice according to the present embodiment. a; (cultivation of Koshihikari in Kagoshima Prefectural agricultural development Center field control group) shows the relationship.

  Further, the line graph indicated by Δ indicates the occurrence ratio of predicted white immature grains with respect to the sample grains M determined to be milky white grains and white dead rice by a grain discriminator (not shown).

On the other hand, FIG. 14 shows the results in the shading area (2 to 22 days after heading and covered with black chilled cold straw with a shading rate of 50%: Koshihikari cultivated in Kagoshima Agricultural Research Center) The relationship between the percentage of brown rice determined to be milky white and white dead rice and the predicted white immature grain and the number of days before harvesting with a grain discriminator (shaded area; 2 to 22 days after heading) And a black chilled ice cream with a light shielding rate of 50%). The line graph which shows the generation | occurrence | production ratio of the prediction white immature grain calculated | required similarly to the case shown in FIG.

  In FIGS. 13 and 14, the lines indicating the vertical widths of the symbols ● and Δ represent standard errors.

13 and 14, from 10 days before harvesting, the white immaturity occurrence rate by “brown rice cross-sectional observation” hardly changes (around 10% in the example shown in FIG. 13 and around 30% in the example shown in FIG. 14) . By observing the cross-sections of glutinous rice and brown rice, it can be seen that the generation ratio of white immature grains at the time of harvesting can be predicted more accurately than the brown rice appearance inspection using a grain discriminator.

  According to the embodiment described above, the grain cutter 1 having the above-described configuration, the harvested rice grain arranging unit 41, the color image scanner which is the imaging unit 51, the computer device 61, the printer 81, or the display unit 64 are used to simplify the process. The ambiguity of the prediction that has been caused by the appearance of the determination of the incidence by brown rice quality before harvesting by the operation alone, the cost and time that have been caused by the appraiser, Reduces individual differences in analysis results without incurring costs, and improves measurement accuracy for predicted sizing, predicted white immature grains, and undetermined grains for eight pre-specified classification categories of rice grains to be harvested It is possible to realize a harvest quality prediction system and a harvest quality prediction method that use a cross-sectional image of the planned harvest rice.

  The harvest-time quality prediction system according to the present invention can be widely applied to objective confirmation of rice that is covered by agricultural mutual aid at the time of harvest, and is used to sort ginger at the time of receipt at a rice dry drying facility. It is also possible to contribute to the improvement of the added value of the product.

DESCRIPTION OF SYMBOLS 1 Grain cutter 2 Apparatus main body 3 Flat part 4 Side wall part 5 Guide rod 6 Rack 7 Moving body 8 Pinion shaft 9 Operation handle 10 Cutting blade holder 11 Screw 12 Cutting blade 13 Screw 14 Safety cover 21 Grain support plate 21a Upper surface part DESCRIPTION OF SYMBOLS 22 Hole 23 Alignment aid 23a Flat part 24 Alignment auxiliary hole 25 Side wall part 31 Transparent body 31a Flat part 32 Side wall part 41 Harvest rice grain arrangement means 51 Imaging means 52 Imaging surface 61 Computer apparatus 62 Program memory 63 Control part 64 Display part 65 Printer Interface 66 Interface 67 Keyboard 68 Classification Category Information Storage Means 69 Image Analysis Processing Means 70 Calculation Means 71 Storage Means 72 Output Information Creation Means 81 Printer M Sample Grain Ma Cut Surface

Claims (4)

A grain cutter that cuts the body of a large number of rice grains to be harvested that are collected from the field on the date of harvest prior to the actual harvest date and stored one by one in each hole of the grain support plate;
The planned harvesting grain arrangement means for arranging the cut surfaces of the sample grains of a large number of planned harvested rice grains cut by the grain cutter,
Imaging means for scanning each cut surface of a plurality of sample grains of the planned harvest rice arranged by the planned harvest grain array means to obtain a cut surface image of each grain;
Image analysis that analyzes the cloudy part of the endosperm part in the cut surface image of each grain acquired by the imaging means with reference to the classification category information designated in advance, and classifies it into predicted grain size, predicted white immature grain, and undecided grain Processing means, analysis of the cloudy portion of the endosperm portion of the cut surface image of each grain that has been treated, said predicted sizing, predicted white immature grain, undetermined grain, and predicted sizing, predicted white immature grain, undetermined grain Based on the number information, the predicted white immature grain generation rate on the harvesting date in a certain period close to the actual harvest date when the predicted white immature grain generation rate hardly changes corresponds to the number of holes in the grain support plate. Calculating means for calculating the number of grains to be denominator and calculating by using the number of predicted white immature grains as a numerator,
Have
Predicting the occurrence ratio of white immature grains on the harvest date based on the calculated occurrence ratio of predicted white immature grains on a collection period in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes The harvest quality prediction system using the cross-sectional image of the planned harvesting rice, which is configured as described above. A grain cutter that cuts the body of a large number of sample grains of brown rice or brown rice that are collected from the field on the day of harvest prior to the actual harvest date and housed in each hole of the grain support plate. When,
The planned harvesting grain arrangement means for arranging the cut surfaces of the sample grains of a large number of planned harvested rice grains cut by the grain cutter,
Imaging means for scanning each cut surface of a plurality of sample grains of the planned harvest rice arranged by the planned harvest grain array means to obtain a cut surface image of each grain;
Analyzing the cloudy part of the endosperm part in the cut surface image of each grain acquired by the imaging means with reference to eight kinds of classification category information designated in advance , and classifying into predicted sizing, predicted white immature grain, and undetermined grain Image analysis processing means, and the predicted sizing, predicted white immature grain, analysis result of the cloudy part of the endosperm portion in the cut surface image of each grain determined as undetermined grain, and predicted sizing, predicted white immature grain, Based on the information on the number of judged grains, the occurrence ratio of the predicted white immature grains on the collection date in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes is calculated for each hole of the grain support plate. The calculation means for calculating the number of grains corresponding to the number as a denominator and the calculation using the number of predicted white immature grains as the numerator, and the storage means for storing the imaging means, the image analysis processing means, and the calculation means Based on the calculation result of the calculation means. Output information generation means for generating a graph information or schedule information on the occurrence rate of white immature grains, a computer device having,
Output means for outputting the processing result of the image analysis processing means and output information creation means as an image, a graph or a table;
Have
Predicting the occurrence ratio of white immature grains on the harvest date based on the calculated occurrence ratio of predicted white immature grains on a collection period in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes It is a quality prediction system at harvest time using a cross-sectional image of the planned harvest rice,
The eight types of classification category information designated in advance are pre-matured rice grains to be harvested, grains whose cut surfaces are totally transparent are set as predictive sized particles, and the grains near the center of the cut surfaces are white and circular. Predicted white immature grains that will become milky grains, grains that are clouded in the vicinity of the center of the cut surface, and those that have already become transparent at the periphery, are predicted white immature grains that become milky or heart white grains. Predicted white immature grains that have a band-like cloudy part at the center as a core white grain, and prediction that a cloudy grain from the back side to the center as well as the outer circumference of the cut surface becomes a milky white or heart white grain When the white immature grain is white and only the backside or ventral periphery of the cut surface is cloudy, the white immature grain is expected to be a white or abdomen white grain, and the cut surface is a milky pasty or whole Grain that is milky white and opaque, is considered as undetermined grain, The section is clarified by starch, the grains peripheral portion is clouded with undetermined grain,
Harvest quality prediction system using cross-sectional images of planned harvest rice. A grain cutter that cuts the body of a large number of sample grains of brown rice or brown rice that are collected from the field on the day of harvest prior to the actual harvest date and housed in each hole of the grain support plate. When,
The planned harvesting grain arrangement means for arranging the cut surfaces of the sample grains of a large number of planned harvested rice grains cut by the grain cutter,
An image scanner that scans each cut surface of a plurality of sample grains of the planned harvest rice arranged by the planned harvest grain array means to obtain a cut surface image of each grain;
Analyzing the cloudy part of the endosperm part in the cut surface image of each grain acquired by the image scanner with reference to eight kinds of classification category information designated in advance , and classifying them into predicted grain size, predicted white immature grain, and undecided grain Image analysis processing means, and the predicted sizing, predicted white immature grain, analysis result of the cloudy part of the endosperm portion of the cut surface image of each grain determined as undetermined grain, and predicted sizing, predicted white immature grain, Based on the information on the number of judged grains, the occurrence ratio of the predicted white immature grains on the collection date in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes is calculated for each hole of the grain support plate. The calculation means for calculating the number of grains corresponding to the number as a denominator and the calculation using the number of predicted white immature grains as the numerator, and the storage means for storing the imaging means, the image analysis processing means, and the calculation means And prediction based on the calculation result of the calculation means Output information generation means for generating a graph information or schedule information on the occurrence ratio of immature grains, a computer device having,
A printer or a display unit for outputting the processing result of the image analysis processing means and the output information creation means as an image, a graph or a table;
Have
Predicting the occurrence ratio of white immature grains on the harvest date based on the calculated occurrence ratio of predicted white immature grains on a collection period in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes It is a quality prediction system at harvest time using a cross-sectional image of the planned harvest rice,
The eight types of classification category information designated in advance are pre-matured rice grains to be harvested, grains whose cut surfaces are totally transparent are set as predictive sized particles, and the grains near the center of the cut surfaces are white and circular. Predicted white immature grains that will become milky grains, grains that are clouded in the vicinity of the center of the cut surface, and those that have already become transparent at the periphery, are predicted white immature grains that become milky or heart white grains. Predicted white immature grains that have a band-like cloudy part at the center as a core white grain, and prediction that a cloudy grain from the back side to the center as well as the outer circumference of the cut surface becomes a milky white or heart white grain When the white immature grain is white and only the backside or ventral periphery of the cut surface is cloudy, the white immature grain is expected to be a white or abdomen white grain, and the cut surface is a milky pasty or whole Grain that is milky white and opaque, is considered as undetermined grain, The section is clarified by starch, the grains peripheral portion is clouded with undetermined grain,
Harvest quality prediction system using cross-sectional images of planned harvest rice. Obtaining a large number of sample grains of rice to be harvested consisting of brown rice or glutinous rice collected from the field on the date of harvest prior to the actual harvest date and stored one by one in each hole of the grain support plate;
A step of cutting the body of a sample grain of a large number of planned rice grains consisting of brown rice or glutinous rice with a grain cutter;
Arranging each cut surface of the sample grains of a large number of planned harvested rice grains cut by the grain cutter by the planned harvesting grain arranging means in a state of being exposed; and
Scanning each cut surface of a plurality of sample grains of the planned harvest rice grains arranged by the planned harvest grain array means with an image scanner to obtain a cut surface image of each grain;
Using a computer device, the cloudy part of the endosperm part in the cut surface image of each grain obtained by the image scanner is the rice grain to be harvested before maturation, and the grain whose entire cut surface has been made transparent is predicted to be sized , The white circular ring around the center of the cut surface is the predicted white immature grain that becomes milky white grain, and the grain that has become cloudy near the center of the cut surface and has already become transparent is the milky white grain Alternatively, the predicted white immature grain that will be a heart white grain, and the grain that has a band-like cloudiness at the center of the cut surface will be the predicted white immature grain that will be a heart white grain, and from the back side to the center in addition to the outer periphery of the cut surface Predicted white immature grains that become milky white grains or heart white grains when they are cloudy, and are predicted to be white or belly white grains when only the backside or ventral periphery of the cut surface is cloudy White immature grains with a milk-like pasty or whole A particle that is opacified in a milky white and undetermined grain, clarified by the center of the cut surface starch, refers to the eight classification category information specified in advance to undetermined grain grain outer peripheral portion is clouded The process of classifying each grain into predicted sizing, predicted white immature grain, and undecided grain, the predicted sizing, predicted white immature grain, and the endosperm portion of the cut surface image of each grain determined as undetermined grain Based on the analysis result of the cloudy part and the predicted sizing, predicted white immature grains, and information on the number of undecided grains, the collection date in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes A process of calculating the generation ratio of predicted white immature grains by calculation using the number of grains corresponding to the number of each hole of the grain support plate as a denominator and the number of predicted white immature grains as a numerator, the result of this calculation process Graph information on the proportion of white immature grains predicted or And performing processing for creating a table information,
Outputting the cut surface image of each grain, the graph or the table to a printer or a display unit;
Including
Predicting the occurrence ratio of white immature grains on the harvest date based on the calculated occurrence ratio of predicted white immature grains on a collection period in a certain period close to the actual harvest date where the occurrence ratio of the predicted white immature grains hardly changes The quality prediction method at the time of harvest using the cross-sectional image of the rice planned to be harvested characterized by being configured as described above.

Images