JP5572963B2 - Quality measuring device - Google Patents

Description

  The present invention relates to a quality measuring apparatus for measuring and selecting the quality of a measurement object conveyed on a conveyor online, and more specifically, the quality varies depending on the direction, shape, size, color, part, etc. The present invention relates to a quality measuring apparatus suitable for measuring the quality of an object to be measured, for example, food such as fruits and vegetables, fish and meat.

There is a quality measurement device of diffuse transmission type or diffuse reflection type that measures the amount of light after irradiating the measurement object from the light source and splitting the transmitted light or reflected light. The transmission light of the light irradiated from the light source to the measurement object to be measured is spectrally separated by the diffraction grating, the spectrum of the required wavelength is measured by the line sensor, and the measurement value is subjected to arithmetic processing and waveform analysis by the signal processing and control device. In addition, there is one that measures the quality of the measurement object online by performing arithmetic processing using a calibration curve for evaluating quality (see, for example, Patent Document 1).
Japanese Unexamined Patent Publication No. 7-229840

  FIG. 9 shows the change in the received light intensity (the intensity of transmitted light from the onion received by the light receiver) depending on the direction of the bud when the measurement object is an onion, and the bud is on the upper side (FIG. 10 ( a)) and the onion V in the front direction (see FIG. 10B) in the conveying direction (see arrow A in the figure), the buds are on the right side in the conveying direction (see FIG. 10). 10 (c)) and the left side (see FIG. 10 (d)), the received light intensity is greatly reduced particularly in the long wavelength region.

  This phenomenon is caused by densely packed onion buds and tissues near the roots, and the quality is measured by a conventional quality measuring device for a measurement object having a different light receiving intensity depending on the direction. Then, since there is a direction of the measurement object whose quality measurement result does not match the actual quality, the discrimination rate is lowered.

  In addition, when quality measurement is performed on a measurement object having an obstacle near the surface by a conventional quality measuring device, information on the obstacle near the surface reflects or reflects light through the measurement object depending on the position of the obstacle. The discriminating rate decreases because the light is not detected as a failure because it is not included in the light, or the light is absorbed by discoloration near the surface and the amount of transmitted light itself decreases and the feature of the failure cannot be detected.

  In this way, when the quality measurement of a measurement object whose quality varies depending on its direction and part (surface obstacle part) or its shape, size, color, etc., the discrimination rate is There is a problem of lowering.

  Therefore, in view of the above situation, the present invention intends to solve the problem of improving the discrimination rate in the quality measurement of a measurement object whose quality varies depending on the direction, shape, size, color, region, etc. It is in providing a quality measuring device that can be used.

In order to solve the above problems, a quality measuring device according to the present invention includes a transport unit that transports a measurement object, a projector that irradiates light to the measurement object transported by the transport unit, A spectroscope that splits transmitted light or reflected light and a light receiver that detects a spectral light quantity of a required wavelength, and a calculation value and a waveform analysis of the detection value of the spectral light quantity detector, and A quality for measuring and selecting the quality of the measurement object online, comprising a signal processing unit that performs arithmetic processing using a calibration curve for evaluating the quality of the measurement object and a quality measurement unit comprising a control device A measuring apparatus, which is installed on the upstream side of the quality measuring unit, images the measuring object, and uses image data of the measuring object imaged by the imaging unit. direction Shape, size, and an image processing apparatus for obtaining a useful image analysis information is analyzed based on the image data such as color or site, by the quality measuring unit when estimating the quality of the measurement object The image analysis information is added to an explanatory variable when the calibration curve is created by multivariate analysis using the image analysis information.

Here, it is preferable to exclude the measurement object having a surface obstacle of a predetermined range or more as a defective product.

In addition, in order to solve the above problems, the quality measuring apparatus according to the present invention includes a transport unit that transports a measurement target, a projector that irradiates light onto the measurement target transported by the transport unit, and the measurement target. A spectroscope that separates transmitted light or reflected light from the light source, a light receiver that includes a spectral light amount detector that detects a spectral light amount of a required wavelength, and arithmetic processing and waveform analysis of the detection value of the spectral light amount detector In order to measure and select the quality of the measurement object online, comprising a quality measurement unit comprising a signal processing and control device that performs arithmetic processing using a calibration curve for evaluating the quality of the measurement object The quality measurement apparatus according to claim 1, wherein the measurement target is installed on the upstream side of the quality measurement unit, and the measurement target is obtained by using the imaging unit that images the measurement target, and the image data of the measurement target captured by the imaging unit. object Direction, shape, size, and an image processing apparatus for obtaining a useful image analysis information is analyzed based on the image data such as color or sites, estimates the quality of the measurement target by the quality measuring unit At the time, using the image analysis information, a plurality of calibration curves are created corresponding to the image analysis information in which the quality of the measurement object varies, and the quality is determined from the image analysis information at the time of actual measurement. The measuring unit selects a calibration curve optimal for quality measurement and estimates the quality of the measurement object.

Here, when estimating the quality of the measurement object by the quality measurement unit, using the image analysis information, the measurement site to be evaluated of the measurement object is determined, the projector and the projector A plurality of light receiver sets are juxtaposed in the transport direction, the measurement site is selected by the quality measurement unit from the image analysis information at the time of actual measurement, and the projector and the light receiver set located at the measurement site It is preferable to estimate the quality of the measurement object based on the above information .

According to the quality measurement apparatus of the present invention, a transport unit that transports a measurement object, a projector that irradiates light to the measurement object transported by the transport unit, and transmitted light or reflected light from the measurement object. And a light receiver comprising a spectroscopic light amount detector for detecting a spectroscopic light amount detector for detecting a spectral light amount of a required wavelength, and calculating and analyzing the waveform of the detection value of the spectroscopic light amount detector, and the quality of the measurement object A quality measuring device for measuring and selecting the quality of the measurement object online, comprising a signal processing unit that performs arithmetic processing using a calibration curve for evaluating The direction, shape, and size of the measurement object using the imaging unit that is installed upstream of the quality measurement unit and that images the measurement object and the image data of the measurement object that is imaged by the imaging unit Well, Or an image processing apparatus for obtaining a useful image analysis information is analyzed based on the image data of the sites and the like, in estimating the quality of the measurement target by the quality measuring unit, the image analysis information When the calibration curve is created by multivariate analysis, the image analysis information is added to the explanatory variable, so the measurement object whose quality varies depending on its direction, shape, size, color, part, etc. In the quality measurement, quality measurement can be performed using useful image analysis information such as the direction, shape, size, color, or part of the measurement object, so that the discrimination rate can be improved.

In addition, for example, when the measurement object is an onion, as the image analysis information, direction information that has different values depending on whether the onion bud is on the light receiving side or the light source side or otherwise is used. Create a calibration curve to evaluate the quality of onions, taking into account the reduction in the amount of light when light passes through the buds and roots where the tissue is densely packed, by using direction information as an explanatory variable. Can do.

  Therefore, it is possible to improve the discrimination rate of an object to be measured, such as an onion, whose received light intensity varies depending on its direction.

In addition to the above-described effects, if the measurement object whose surface obstruction is greater than or equal to a predetermined range is excluded as a defective product, surface obstruction information is not included in transmitted light and is not detected as obstruction, or discoloration near the surface As a result, it is possible to reduce the number of cases where the light is absorbed and the transmitted light amount itself is reduced and the feature of the failure cannot be detected, so that the discrimination rate can be further improved.

Furthermore , according to the quality measuring apparatus according to the present invention, a transport unit that transports the measurement object, a projector that irradiates light to the measurement object transported by the transport unit, transmitted light from the measurement object, or A photoreceiver comprising a spectroscope that divides the reflected light and a spectroscopic light amount detector that detects a spectroscopic light amount of a required wavelength, and performs calculation processing and waveform analysis on the detection value of the spectroscopic light amount detector, and the measurement object A quality measurement device for measuring and selecting the quality of the measurement object online, comprising a signal processing unit that performs arithmetic processing using a calibration curve for evaluating the quality of the product and a quality measurement unit comprising a control device An image capturing unit configured to image the measurement target, which is installed on the upstream side of the quality measurement unit, and the direction and shape of the measurement target using image data of the measurement target captured by the image capturing unit. , Can of, and an image processing apparatus for obtaining a useful image analysis information is analyzed based on the image data such as color or site, in estimating the quality of the measurement target by the quality measuring unit, wherein Using the image analysis information, a plurality of calibration curves are created corresponding to the image analysis information in which the quality of the measurement object fluctuates, and the quality measurement unit determines the quality from the image analysis information at the time of actual measurement. Since the quality of the measurement object is estimated by selecting a calibration curve that is optimal for the measurement, the quality of the measurement object varies depending on its direction, shape, size, color, part, etc. Since quality measurement can be performed using useful image analysis information such as the direction, shape, size, color, or part of the image, the discrimination rate can be improved.
In addition, since the quality measurement can be performed using the optimal calibration curve based on the image analysis information, the discrimination rate can be improved.

Furthermore, when estimating the quality of the measurement object by the quality measurement unit, the image analysis information is used to determine a measurement site to be evaluated of the measurement object, and the projector and the projector A plurality of light receiver sets are juxtaposed in the transport direction, the measurement site is selected by the quality measurement unit from the image analysis information at the time of actual measurement, and the projector and the light receiver set located at the measurement site When the quality of the measurement object is estimated based on the information of the measurement object, in addition to the effect, it is desired to evaluate by selecting and measuring a measurement site suitable for the quality to be evaluated in the measurement object using the image analysis information Efficient quality measurement can be performed while improving the accuracy of information.

  Next, embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to the embodiments shown in the accompanying drawings, and includes all the embodiments that satisfy the requirements described in the claims. It is a waste. In addition, in this specification, the conveyance direction (refer arrow A in the figure) of the onion V shown as an example of the measuring object whose quality is measured online by the quality measuring device of the present invention is the front, and the opposite side is the rear. And the left and right are forward-facing, and the view from the left is the front view.

  1 to 7 are explanatory views of a configuration of a quality measuring apparatus according to an embodiment of the present invention. FIG. 1 is a perspective view, FIG. 2 is a perspective view showing a state in which a cover 7 is removed in FIG. 4 is a front view, FIG. 4 is an enlarged front view showing the imaging unit 2, FIG. 5 is a plan view, FIG. 6 is a view seen from the rear, and FIG. 7 is a configuration of the size measuring unit 3 and the quality measuring unit 4. It is a schematic plan view which shows.

  In FIGS. 1 to 3, the front and rear of the conveyance belt 1 </ b> A constituting the conveyance unit 1 are shown cut. However, the conveyance belt 1 </ b> A is endless and is positioned before and after the base 6. The drive pulley and the driven pulley (not shown) supported so as to be rotatable around the left and right axis are wound around the motor, and the motor connected to the drive pulley is rotated at a constant speed so that the drive pulley and the driven pulley are fixed. Since the upper portion of the endless transport belt 1A moves at a constant speed in the transport direction (see arrow A in the figure), the onions V,... Mounted on the transport belt 1A are rotated. It is transported at a constant speed in the transport direction.

  The details shown in FIG. 2 are covered by support frames 20A, 20B, and 20C fixed on the base 6 so as to cover the imaging unit 2, the size measuring unit 3, and the quality measuring unit 4 described later. In the state shown in FIG. 1 to which the cover 7 is attached, the strip-shaped light-shielding curtains 9 and 9 are attached to the front and rear openings 8 and 8 of the cover 7, so that intrusion of external light into the cover 7 is suppressed. .

  As shown in FIG. 2, FIG. 3 and FIG. 7, the quality measuring apparatus according to the embodiment of the present invention has been transported by the transport unit 1 that transports the onions V,. An on-line imaging unit 2 that images the onion V, a size measuring unit 3 that is installed on the downstream side of the imaging unit 2 and that measures the size of the onion V online, and is arranged on the downstream side of the size measuring unit 3. The projector 21E that irradiates the onion V with light of a predetermined wavelength (for example, near-infrared light), the light receiver 21R that spectrally processes the transmitted light from the onion V, and the image data of the imaging unit 2 are processed as described later. At the same time, the accumulation time of the line sensor 32 which is a spectral light quantity detector is adjusted in accordance with the size measurement value of the onion V by the size measuring unit 3, the measurement value of the line sensor 32 is subjected to arithmetic processing and waveform analysis, and the quality of the onion V is determined. Review Performing arithmetic processing using the calibration curve (discriminant) for, and a quality measuring unit 4 consisting of the image-signal processing and control unit 5, which is furnished to the housing 10.

  The imaging unit 2 and the size measuring unit 3 only need to be positioned upstream of the light projector 21E and the light receiver 21R of the quality measuring unit 4, and the imaging unit 2 is installed downstream of the size measuring unit 3. May be.

  Next, the projector 21E, the light receiver 21R, and the size measuring unit 3 constituting the quality measuring unit 4 will be described with reference to FIG.

  The projector 21E is installed, for example, on the left side of the transport belt 1A along the transport path of the onions V,..., A light-shielding light source case 22, a halogen lamp 23 that is a light source attached to the side surface of the light source case 22, A reflector 24 fixed in the light source case 22 that reflects light from the halogen lamp 23, a condenser lens 25 installed on the inner side (left side) of the light projection port 22A formed on the right side of the light source case 22, and The shutter includes a light source shutter 26 and the like which are opened and closed by an actuator such as a solenoid (not shown) to prevent unnecessary light from being irradiated when the measurement is stopped.

  Further, the light receiver 21R is installed along the transport path of the onions V,... On the right side of the transport belt 1A so as to face the projector 21E, and is formed on the left side of the light shielding case 27 and the light shielding case 27. The condensing lens 28 installed on the inner side (right side) of the light passage opening 27A, the visible light cut filter 29A, the slit 29B, and the measurement light installed sequentially from the condensing lens 28 to the right side (FIG. 7). The shutter driving actuator that switches between the state in which the dark current of the line sensor 32 is measured by the offset acquisition shutter (black shutter) 30A and the state in which the reference light is received by the ND filter (neutral density filter) 30B. 30, and is split by the flat field concave diffraction grating 31 and the diffraction grating 31. Read collectively the light quantity of each spectral that focused on in, consisting of the line sensor 32 and the like of the charge accumulation method is a multi-channel spectral intensity detector.

  When the quality of the onion V is measured, when the light source shutter 26 is opened, the light condensed by the condenser lens 25 of the light projector 21E is conveyed by the conveying unit 1 from the light projection port 22A of the light source case 22. The transmitted light from the onion V that has been irradiated to V reaches the inside from the light passage opening 27A of the light shielding case 27 of the light receiver 21R, and the transmitted light converged at the position of the slit 29B by the condenser lens 28 is reflected by the diffraction grating 31. Spectroscopy is performed, and the spectrum of the required wavelength is measured by the line sensor 32.

  The measured values measured in this manner are subjected to arithmetic processing and waveform analysis by the image / signal processing and control device 5 (see FIGS. 1 to 3), and are calculated using a calibration curve for evaluating the quality of the onion V. Processed and the quality of the onion V is measured.

  The sensor used in the size measuring unit 3 installed on the upstream side of the projector 21E and the light receiver 21R of the quality measuring unit 4 is, for example, a transmission type photosensor including a light emitting element 18 and a light receiving element 19, and the sensor is in a detection state. The size of the onion V is calculated by the image / signal processing and control device 5 from the time from when it becomes the non-detection state and the conveying speed of the onion V, and based on this calculated value, the light receiver The image / signal processing and control device 5 controls the shutter driving actuator 30 so that the timing of opening the shutter in 21R (timing of measurement start) is matched with the arrival of the onion V being conveyed.

  Next, the configuration of the imaging unit 2 and the image processing of the image data captured by the imaging unit 2 will be described with reference to FIGS.

  The imaging unit 2 is fixed to each of the front and rear, left and right sides of the rectangular frame on the lower side of the camera 11 that images the onion V conveyed by the conveying unit 1 from above, and faces the onion V. In this way, the lighting 12 illuminates the onion V from the front, rear, left and right, and the photo sensor 13 for taking the shutter timing of the camera 11.

  In addition, a polarization filter 15 is attached to the camera 11 below the lens 14, and a polarization filter 17 is attached to the illumination 12 below the LED unit 16.

  The onion V image data captured online by the imaging unit 2 is sent to the image / signal processing and control device 5, and the image / signal processing and control device 5 detects the direction of the onion V. Image processing and image processing for extracting the surface disorder of the onion V are performed.

  That is, the onion V image data picked up by the image pickup unit 2 is extracted as, for example, a monochrome binary image, the contour of the onion V is extracted, an ellipse is curve-fitted to the contour, and the long axis direction of the ellipse is obtained. The direction of V (the direction of buds) can be determined.

  For example, it is assumed that the distance from the center of the ellipse to the circumference of the ellipse and the distance to the contour of the onion V are obtained in all directions, and buds are present in the direction in which the difference between these distances is maximized.

  When the onion V buds are facing upward, the error between the ellipse and the contour is reduced, and it is assumed that the onion V buds are facing upward when this error is equal to or less than a predetermined value.

  In addition, the image data of the onion V imaged by the imaging unit 2 is an HSV composed of hue (H), saturation (S), and lightness (V) that enables discrimination close to human vision from, for example, the RGB color system. The lightness (V) of the portion having no surface hindrance (rot) is 60 to 255, while the lightness (V) of the surface hindrance portion is higher than the lightness of the portion having no surface hindrance (rotation). Since it is also dark and is about 60 to 150, it is possible to extract a surface obstacle portion by paying attention to the difference in brightness.

  It should be noted that the degree of surface obstruction is determined from the ratio of the area of the surface obstruction portion extracted in this way to the total area (for example, the surface obstruction level is 0 if there is no surface obstruction, and the surface obstruction levels 1-3 depending on the degree of surface obstruction. If the degree is equal to or greater than a predetermined value, the defective product may be determined as a defective product.

  Next, an example of creating a calibration curve including the above-described onion V direction (bud direction) will be described.

  When the direction of the bud is the right side (light receiving side) or the left side (light source side) as shown in FIG. 10C and FIG. 10D, light is transmitted through the densely packed tissue near the bud and the root. For this reason, the received light intensity (transmitted light amount) decreases.

  Therefore, the secondary differential spectrum of the spectrum measured by the line sensor 32, which has been subjected to the secondary differential process to remove the direct current portion due to the change in the size of the onion V, is used as the explanatory variable, and further, the above-described image When the onion V direction (bud direction) obtained by the processing is the right side (light receiving side) or the left side (light source side), the case where the received light intensity decreases as described above is set to 2, for example, and the other direction is set to, for example, 1 and such direction information is added to the explanatory variable.

  Then, the onion V internal failure (internal rot) is turned onion according to a predetermined rot evaluation template (for example, those having no internal failure are set to internal failure level 1 and internal failure levels 2 to 5 depending on the degree of internal failure). A price is measured by observing the cross section of V, and a calibration curve is created by performing multivariate analysis using this value as an objective variable.

  For example, 100 onions V,..., When the bud direction is the upper side (see FIG. 10A), the secondary differential spectrum of the spectrum and the direction information 1, and the bud direction is the front side ( The secondary differential spectrum of the spectrum and the direction information 1 as shown in FIG. 10 (b)), and the secondary of the spectrum when the direction of the bud is on the right side (see FIG. 10 (c)). The differential spectrum and the direction information 2 and the secondary differential spectrum of the spectrum when the bud direction is set to the left side (see FIG. 10D) and the direction information 2 are explanatory variables, and these 100 onions. A calibration curve may be created by performing multivariate analysis with V,... Internal failure levels 1 to 5 as objective variables.

  When the direction information is used as an explanatory variable when creating a calibration curve by multivariate analysis in this way (Example 1), and the same calibration curve as in Example 1 is used, and surface defects are extracted by the above-described image processing. When the surface defect is excluded as a defective product (Example 2), and when creating a calibration curve by multivariate analysis corresponding to the prior art, direction information is not used as an explanatory variable. FIG. 8 is a table showing the results of comparing the discrimination rates of the quality measuring devices for each threshold value for determining the presence or absence of a failure in the case of (comparative example).

  From FIG. 8, it can be seen that when the comparison is made at the highest discrimination rate, the discrimination rate is 88.0% in Example 1 and 92.4% in Example 2 as compared to 85.8% in the comparative example.

  That is, according to the configuration of the first embodiment, using the direction of the onion V obtained by the image processing device (image / signal processing and control device 5), the buds of the onion V are on the right side (light receiving side) or the left side (light source side). ) And direction information with different values depending on other cases are also used as explanatory variables, taking into account the decrease in light quantity when light passes through the buds and roots where tissues are densely packed. Since a calibration curve for evaluating the quality of the onion V can be created, the discrimination rate can be improved.

  Further, according to the configuration of the second embodiment, the onion V from which the surface failure is extracted by the image processing device (image / signal processing and control device 5) is excluded as a defective product, and the onion V is not shipped. Therefore, it is possible to reduce the number of cases where surface obstruction information is not included in transmitted light and is not detected as obstruction, or when light is absorbed due to discoloration near the surface and the amount of transmitted light itself is reduced and the characteristics of the obstruction cannot be detected. Therefore, the discrimination rate can be improved.

  Here, the image analysis information obtained by analyzing the image data of the measurement object imaged by the imaging unit 2 by the image processing device (image / signal processing and control device 5) is limited to the direction and the part (surface obstacle part). It is not intended to include all useful information that is analyzed based on the image data of the imaged measurement object, such as the shape, size, and color.

  By using such useful image analysis information when the quality measurement unit 4 estimates the quality of the measurement object, the discrimination rate can be improved.

  In the above description, the configuration in which the image / signal processing and control device 5 is housed in the housing 10 has been described. However, the signal processing and control device and the image processing device may be separated, and the image processing device may be an imaging unit. You may install near 2.

  Whether there are buds, roots, etc. between the projector 21E and the light receiver 21R (hereinafter referred to as on the optical axis) as in the measurement object having buds and roots such as the onion V shown in the quality measurement example above. For measurement objects with different received light intensities depending on whether or not buds and roots are on the optical axis, two near-infrared spectra are measured and two calibration curves are created. From the image analysis information (direction of the measurement object) obtained by analyzing the image data of the measurement object at the time of analysis by the image processing device (image / signal processing and control device 5), the quality measurement unit 4 is optimal for quality measurement. One calibration curve may be selected for quality measurement.

  Alternatively, in a measurement object whose thickness varies depending on the direction (posture), a calibration curve is created for each thickness by changing the thickness on the optical axis of the measurement object in stages, and in the actual measurement From the image analysis information (direction of the measurement object) obtained by analyzing the image data of the measurement object by the image processing device (image / signal processing and control device 5), the quality measurement unit 4 generates a calibration curve optimal for quality measurement. You may select and perform quality measurement.

  In this way, a plurality of calibration curves are created corresponding to the image analysis information in which the quality of the measurement object varies, and the calibration curve optimal for quality measurement is obtained from the image analysis information at the time of actual measurement by the quality measurement unit 4. By selecting and estimating the quality of the measurement object, quality measurement can be performed using an optimal calibration curve based on the image analysis information, so that the discrimination rate can be improved.

  Next, an example in which the measurement object is not a fruit and vegetable such as the onion V but a fish will be described.

  For example, in tuna and bonito, fat is concentrated in the abdomen. When eating raw, it is better to have more fat, but when processing into bonito, less is better. When evaluating these fat contents, the value of a part of fat is also important, but it is also important how many parts more than a certain fat are present. Such fish heads and tails are not subject to evaluation and must be excluded from measurement.

  Here, tuna and bonito fish bodies have a spindle shape, and the position where the width is the largest is not affected by the head or tail. Or you may make it analyze the color of a fish body and select the location more suitable for evaluation.

  In such a measurement object, as the image analysis information obtained by analyzing the image data by the image processing device (image / signal processing and control device 5), the position selected according to the shape of the fish as described above or as described above. The position selected from the color of the fish body is selected as a measurement site, for example, a plurality of sets of light projectors 21E and light receivers 21R are juxtaposed in the front-rear direction, and the fish as the measurement object is passed between them, It is possible to evaluate how much high fat is present based on the information of the set of the light projector 21E and the light receiver 21R located at the measurement site.

  In this way, a measurement part to be evaluated for the measurement object is determined, and the quality measurement unit 4 selects the measurement part from the image analysis information in actual measurement to estimate the quality of the measurement object. Therefore, it is possible to select and measure a measurement part suitable for the quality to be evaluated in the measurement object using the image analysis information, and thus it is possible to perform efficient quality measurement while improving the accuracy of the information to be evaluated. .

  When the measurement object is a fish as described above, measurement may be performed by bringing a probe including a light projecting fiber and a light receiving fiber into contact with the measurement site of the fish body.

It is a perspective view of a quality measuring device concerning an embodiment of the invention. It is a perspective view which shows the state which removed the cover in FIG. It is also a front view. It is a front view which expands and shows an imaging part. It is also a plan view. Similarly it is the figure seen from back. It is a schematic plan view which shows the structure of a size measurement part and a quality measurement part. It is a figure which shows the discrimination | determination rate for every determination threshold value in an Example and a comparative example by table | surface. It is a figure which shows the change of the light reception intensity by the direction of the bud of an onion. It is a top view which shows the direction of the bud of an onion, (a) When a bud is an upper side, (b) When a bud is a front side, (c) When a bud is on the right side, (d) is a bud on the left side Shows the case.

A Transport direction V Onion (object to be measured)
DESCRIPTION OF SYMBOLS 1 Conveying part 2 Imaging part 3 Size measuring part 4 Quality measuring part 5 Image / signal processing and control device 11 Camera 12 Illumination 13 Photo sensor 14 Lens 15 Polarizing filter 16 LED unit 17 Polarizing filter 21E Projector 21R Receiver 31 Diffraction grating (spectral) vessel)
32 line sensor (spectral light detector)

Claims (4)

A transport unit that transports the measurement object, a projector that irradiates light onto the measurement object transported by the transport unit, a spectroscope that separates transmitted light or reflected light from the measurement object, and a wavelength of the light that is required A light receiving device comprising a spectral light amount detector for detecting the spectral light amount, and a calculation value and a waveform analysis of the detection value of the spectral light amount detector, and calculation using a calibration curve for evaluating the quality of the measurement object A quality measurement device for measuring and selecting the quality of the measurement object online, comprising a quality measurement unit comprising a signal processing and control device for processing,
An imaging unit installed on the upstream side of the quality measurement unit for imaging the measurement object;
Useful image analysis information analyzed based on the image data such as the direction, shape, size, color, or part of the measurement object using the image data of the measurement object imaged by the imaging unit An image processing device to be obtained,
When the quality measuring unit estimates the quality of the measurement object, the image analysis information is used, and when the calibration curve is created by multivariate analysis, the image analysis information is added to an explanatory variable. A quality measuring device characterized by that.   The quality measuring apparatus according to claim 1, wherein the measurement object having a surface obstacle of a predetermined range or more is excluded as a defective product. A transport unit that transports the measurement object, a projector that irradiates light onto the measurement object transported by the transport unit, a spectroscope that separates transmitted light or reflected light from the measurement object, and a wavelength of the light that is required A light receiving device comprising a spectral light amount detector for detecting the spectral light amount, and a calculation value and a waveform analysis of the detection value of the spectral light amount detector, and calculation using a calibration curve for evaluating the quality of the measurement object A quality measurement device for measuring and selecting the quality of the measurement object online, comprising a quality measurement unit comprising a signal processing and control device for processing,
An imaging unit installed on the upstream side of the quality measurement unit for imaging the measurement object;
Useful image analysis information analyzed based on the image data such as the direction, shape, size, color, or part of the measurement object using the image data of the measurement object imaged by the imaging unit An image processing device to be obtained,
When estimating the quality of the measurement object by the quality measurement unit, the image analysis information is used to create a plurality of calibration curves corresponding to the image analysis information in which the quality of the measurement object fluctuates. A quality measurement apparatus characterized in that the quality measurement unit selects a calibration curve optimal for quality measurement from the image analysis information in actual measurement and estimates the quality of the measurement object. In estimating the quality of the measurement target by the quality measuring unit, by using the image analysis information, along with the previously determined measurement site to be evaluated of the measurement object, the projector and the light receiver set Are arranged side by side in the transport direction, the measurement part is selected by the quality measurement unit from the image analysis information at the time of actual measurement, and the information on the projector and the receiver set located in the measurement part is used. The quality measuring apparatus according to claim 1 or 3, wherein the quality of the measurement object is estimated.

Images