JP2008541007A - Food foreign matter detection device - Google Patents

Abstract

In the foreign matter detection apparatus 10 for detecting foreign matter such as hair and insects mixed in the food 20, the transport conveyor 11 for transporting the food 20 to the detection area A and the food from at least two directions across the detection area A Illumination means 14 and 15 for irradiating illumination light, an imaging camera 12 for imaging the front image of the food from one side of the detection area A, and image processing including binarization processing for the captured front image An image processing means 13 for detecting foreign matter, and the image processing means 13 binarizes the captured table image, and exceeds a preset blackness threshold value among the obtained binary images. After detecting an abnormal pixel, the presence / absence of abnormality of a pixel adjacent to the one pixel is detected, and when the abnormal pixel continuously appears based on the one pixel, it is determined as a foreign object. At this time, when abnormal pixels appear in a continuous line shape, they are determined as fibrous foreign matters such as hair, and when they appear in a continuous area shape, they are determined as area foreign matters such as insects.

Description

  The present invention relates to a foreign matter detection apparatus that detects foreign matter in food, and more particularly to a foreign matter detection device for food that detects foreign matters such as fibrous foreign matter such as hair and area foreign matter such as insects attached to the surface of food.

In various food manufacturing factories, foreign substances may be mixed in food ingredients or in the manufacturing stage, and these foreign substances must be removed from the viewpoint of hygiene management in order to sell them as products. It is requested to do. Various kinds of insects are the most common foreign substances mixed in food, and hair is then cited. These two types of foreign matter account for approximately 50% of the total foreign matter.
Possible causes of hair contamination include cases where hair is lost or mixed in when hair is attached to clothes in the process involving workers, or hair that has already fallen on the floor is wrapped by static electricity. Examples include cases where the material adheres to and mixes through the material, and foreign matter is often mixed in the product cooling stage, the handling stage after placing, the product packaging stage, and the like.
On the other hand, there are cases where insects are mixed into foods, such as cases where they are mixed into food ingredients as carcasses of adults, cases where they are mixed with eggs, or cases where eggs such as eggs and cockroaches that have already been mixed are mixed. Or a case where insects flying from outside the factory are mixed.
Conventionally, in these foreign object inspections, human beings as workers have inspected these foreign object contaminations by visual inspection, but it takes time and has a problem in accuracy.

In recent years, various automated foreign object detection devices for food have been proposed and put into practical use. Examples of such foreign matter detection devices include devices using electromagnetic waves, devices using X-rays, and devices using image processing using visible light. Since the apparatus using electromagnetic waves is applied only to foreign substances that react with electromagnetic waves such as metals, the purpose of use has been limited. An apparatus using X-rays is an apparatus that performs determination based on a color difference on a black and white image that appears when X-rays are transmitted. Currently, generally used X-ray detectors have a large color difference for metals such as iron and are easy to distinguish, but cannot detect foreign substances that are components (proteins) that are almost the same as food, such as hair and insects. It was difficult to apply to proteinaceous foreign matter. In addition, since the X-ray detection apparatus requires an area of 1 mm ² or more, there has been a problem that fibrous foreign matters having elongated shapes such as hair cannot be detected.

On the other hand, devices using image processing with visible light are disclosed in Patent Document 1 (Japanese Patent Laid-Open No. 5-223950), Patent Document 2 (Japanese Patent Laid-Open No. 2001-21499), and Patent Document 3 (Japanese Patent Laid-Open No. 2004-245695). Gazette) and the like.
Patent Document 1 is an apparatus for detecting hair mixed in a packed food. An optical image of the food is picked up by an image pickup device, and the shape of the hair is extracted as a partial feature from data obtained by binarizing the picked-up optical image. In addition, the hair is specifically detected by determining whether or not the extracted data is a linear body having a fineness below a certain value and a length above a certain value.

Further, in Patent Document 2, food such as sake lees and noodle dough is formed into a plate having a certain thickness and is fed in a vertical direction by a roller device. A configuration of an apparatus for detecting foreign matter by acquiring transmitted light and performing image processing is described.
Further, in Patent Document 3, a gap is provided between the carry-in conveyor and the carry-out conveyor, a means for transmitting and illuminating the detection target on one side of the conveyor, and a line sensor provided at a position sandwiching the transmitted illumination means and the conveyor And an apparatus for detecting a foreign object in a detection target by binarizing an image obtained by scanning by a line sensor.
Using these devices, it is possible to detect various types of foreign matters mixed in food by acquiring an optical image of the target food and analyzing the obtained image data.

However, the apparatus described in Patent Document 2 is limited to foods whose detection target can be formed into a plate shape, and is about packed frozen udon, frozen rice, or a molded product having a complicated shape such as dumplings or shumai. Is not applicable. Moreover, in patent document 1, when the shape is not uniform like the said molded product and the foodstuff whose surface unevenness | corrugation is remarkable is a test object, there are many shadow parts on the imaged image, and it is binary. There is a possibility that the shadow portion may be misrecognized as a foreign object when the foreign object is detected by the conversion processing. Furthermore, the apparatus described in Patent Document 3 can be applied only to a detection target that transmits light, and when this apparatus is also applied to a detection target having an uneven surface, There is a problem in that a shadow is generated in imaging, so that a false recognition rate is high and detection accuracy is lowered.
Furthermore, when these conventional foreign object detection devices are applied to food, it is necessary to clear the problem of hygiene, unlike detection objects other than food.
JP-A-5-223950 JP 2001-21499A JP 2004-245695 A

  Accordingly, in view of the above-described problems of the prior art, an object of the present invention is to provide a food foreign matter detection device capable of detecting foreign matter with high accuracy even for a detection target having unevenness. .

Therefore, in order to solve the problem, the present invention provides a foreign object detection device that detects foreign substances mixed in food.
A transport unit configured to transport the food to a foreign object detection region; an illumination unit configured to irradiate the food with illumination light from at least two directions across the detection region; and a front image of the food from one side of the detection region. Imaging means for imaging, and image processing means for performing image processing including binarization processing on the captured table image,
The image processing means binarizes the captured table image, and after detecting one abnormal pixel exceeding a preset threshold value in the obtained binary image, an abnormality of a pixel adjacent to the one pixel is detected. The presence / absence of a foreign object is detected, and when an abnormal pixel continuously appears based on the one pixel, it is determined as a foreign object.

  According to the present invention, it is possible to detect a foreign substance in food that has been difficult to detect with high accuracy. By illuminating the food from at least two directions across the detection area, the shadow caused by the unevenness on the surface of the food is erased, and the mixed foreign matter is raised from the surrounding food. Even a food having a simple shape can detect foreign matter with high accuracy.

In addition, when the abnormal pixel appears in a continuous line shape based on the one pixel, it is determined as a fibrous foreign matter such as hair.
In this way, the continuity of abnormal pixels exceeding the threshold values of the length and thickness related to the preset line shape in the binary image obtained by the binarization processing is detected, and the abnormal pixels appear in a continuous line shape. Only when it is determined that the foreign substance is a foreign substance, only the fibrous substance can be selectively detected.

The abnormal pixel is determined to be an area foreign object when it appears in a continuous area shape based on the one pixel.
Thus, to detect the continuity of the abnormal pixel exceeding the threshold value for the area set in advance in the binary image obtained by binarizing process, this only if shows the area shape anomaly pixels adjacent By determining as a foreign object, it is possible to selectively detect only an area foreign object.
The area foreign matter means a foreign matter having a predetermined width and length, or a foreign matter having a predetermined area or more.

Furthermore, the 1st illumination means located in the said imaging means side among the said illumination means is a means to irradiate visible light, and the 2nd illumination means located in the other side is a means to irradiate infrared light. It is characterized by that.
Thus, by irradiating infrared light with high light transmittance from the opposite side of the image pickup means with the belt in between, it is possible to eliminate most of the shadow caused by the unevenness of the food. Furthermore, it is preferable to provide two or more of the first illumination means and irradiate one side of the food (imaging means side) from two or more different directions.

The first illumination means is a means for irradiating any one of red, blue, green, and white light as a mixed color, and is appropriately selected according to the type of food.
In this way, by using a light source corresponding to food, food can be imaged more clearly, and highly accurate foreign object detection can be performed.

Furthermore, the imaging means is a CCD camera, and the focal length of the lens of the CCD camera is in the range of 500 mm to 900 mm, preferably 800 mm to 900 mm.
This makes it easy to focus the captured image even on foods with significant irregularities or foods with a large distance between the food surface and the bottom of the tray.
Furthermore, the transport means includes a cleaning means for cleaning a portion that comes into contact with the food, and at least any one of the illumination means is configured to be movable, and the illumination means is moved from the transport means during cleaning by the cleaning means. It is characterized in that it is stored in a storage box that is provided separately.
By providing the cleaning means, it is possible to provide a product that is safe and reliable in terms of food hygiene. In addition, the lighting means is stored in a storage box, so that water or cleaning liquid is supplied to the lighting means. Therefore, it is possible to prevent a malfunction from occurring. In addition, it is preferable that the illumination means accommodated in the said storage box is a said 1st illumination means.

The second illuminating means is housed in a sealed storage box formed of a material that transmits infrared light on the side facing the transporting means, and the sealed storage box includes a cooling means for cooling the inside. It is characterized by that.
Accordingly, when the cleaning means is provided, it is possible to prevent the second lighting means from being exposed to a cleaning liquid or water, and to prevent a malfunction, and the second lighting means is located below the gap between the conveyors. In the case of the installed configuration, it can be easily cleaned even if garbage falling from the food adheres.

Further, the conveying means has two conveyor belts arranged in parallel, and the two conveyor belts are configured so that the transport directions of the two conveyor belts are opposite to each other, and food is provided at the end of the conveyor belt. A reversing unit is provided, and the foreign matter on the front surface of the food is detected on one conveyor, and the foreign matter on the back surface of the food is detected on another conveyor.
In this way, by configuring the conveying means with a conveyor belt composed of an outward path and a return path, it is possible to detect foreign matter adhering to both the front and back surfaces of the food with a single detection device. Moreover, since foreign matter can be detected on both the front and back surfaces at once, the inspection time can be shortened.
Furthermore, the foreign substance is a substance mainly composed of a protein. In the present invention, even a foreign substance mainly composed of a protein that could not be detected by a conventional X-ray detection apparatus can be suitably detected.

As described above, according to the present invention, it is possible to detect foreign matters such as hair and insects that have been difficult to detect with high accuracy. In addition, by illuminating from two directions across the detection area, the shadow caused by the unevenness of the food surface can be erased, and the mixed foreign matter can be raised from the surrounding food, so that the shadow is mistakenly recognized as a foreign matter. Can be prevented. In addition, by detecting abnormal continuity exceeding a preset blackness threshold among binary images obtained by binarization processing, it is possible to selectively detect only fibrous substances or area foreign matters. It becomes possible.
Furthermore, by providing a cleaning means for cleaning the conveying means, it is possible to provide a safe and secure product in terms of hygiene management, and when the lighting means can be stored, water, cleaning liquid, etc. are generated due to adhesion. You can prevent problems.
In addition, by providing a transport means composed of a reciprocating path provided with a reversing means, it becomes possible to detect foreign matter adhering to both the front and back surfaces of food with a single detection means, and also shorten the inspection time. I can plan.

Hereinafter, the present invention will be described in detail with reference to the embodiments shown in the drawings. However, the dimensions, materials, shapes, relative arrangements, and the like of the component parts described in this example are not intended to limit the scope of the present invention only to specific examples unless otherwise specified. Only.
The inspection object of this embodiment is food, which includes food alone or food stored in a tray. The type of food is not particularly limited, and examples thereof include white rice, udon, shumai, dumplings, and the like. These works may be in a frozen state, a normal temperature state, or a heated / heat-retained state. Foreign matter to be detected is foreign matter such as hair and insects.
1 is a front view of a foreign object detection apparatus according to an embodiment of the present invention, FIG. 2 is a side view of the foreign object detection apparatus shown in FIG. 1, and FIG. 3 is a schematic front view showing the foreign object detection apparatus provided with cleaning means. 4 is a schematic front view showing a foreign matter detection apparatus according to another embodiment, FIG. 5 is a schematic plan view showing another embodiment of the transfer conveyor, and FIG. 6 is a binary value obtained by the image processing apparatus of this embodiment. It is a figure which shows an image.

A foreign object detection apparatus 10 shown in FIGS. 1 and 2 includes a transport conveyor 11 that transports food to a detection area A, and an imaging camera 12 that captures a front image of the food placed on the transport conveyor 11 in the detection area. And image processing device 13 for detecting foreign matter by image processing the front image of the food obtained by the imaging means, and illumination light on the food in detection area A from two different directions across the conveyor 11 The illumination devices 14 and 15 that irradiate are the main components. The detection area A substantially matches the field of view of the imaging camera 12.
Further, in order to improve the operability, the foreign object detection device 10 includes a monitor 22 on which a table image of food imaged by the imaging camera 12 is displayed, a touch panel 23 and operation keys 24 for operating the device 10. It is preferable to include a lifting handle 25 or the like that can adjust the height of the upper illumination device 14 according to the size of the food to be inspected.

The conveyor 11 has a structure in which a conveyor belt 11a is stretched around a driving roller 11b connected to a motor.
A specific form of the conveyor 11 is shown in FIG. As shown in the figure, the conveyor belt 11a is formed of an infrared light transmissive material, and the lower illumination means 15 is installed under the upper belt in the food detection area A. At this time, the conveyor belt 11a may be formed of a transparent or translucent material, for example, a urethane material or the like. Even if the belt is white, it may be used depending on the food. As another form, as shown in FIG. 4, a plurality of the conveyor belts 11a are arranged in series, a gap is provided at a position corresponding to the food detection area A, and the lower illumination device 15 is installed below the gap. You may do it. In this case, the conveyor belt 11a may not be infrared transmissive, and the material is not particularly limited.

The illumination means includes an upper illumination device 14 disposed above the conveyor 11 and a lower illumination device 15 disposed inside or below the conveyor 11.
The upper illumination device 14 is arranged above the detection area A so as to illuminate the food from both sides, similarly to the installation position of the imaging camera 12. The upper illumination device 14 causes the foreign matter adhering to the food to rise from the surrounding food, thereby creating a reference point for recognition of the presence of the foreign matter. For example, when the foreign substance is hair, the cross-sectional shape is circular, and thus the side surface of the hair is illuminated by such an illumination method so that it can be distinguished from the surrounding food. When the foreign object is an insect, its cross-sectional shape is larger than a certain size, so the side of the insect is illuminated to distinguish it from the surrounding food. With respect to other foreign matters, the surrounding food and foreign matters can be clearly distinguished by adopting an illumination method that conforms to the shape of the foreign matter.
Further, the upper illumination device 14 uses visible light, and is an illumination in which foreign matters appearing on the back of the food from four elements of red, blue and green which are the three primary colors of light and white light mixed with these. Select a condition.

At this time, when the foreign matter is black hair, the foreign matter is present in a portion close to black in the black and white binarized image. First, a target region is set in a range including this portion, and a pixel satisfying this condition is set as a verification target. The same applies to white hair.
When the foreign matter is an insect, the foreign matter is present in a portion close to black in the black binarized image. Therefore, as in the case of hair, a target region is first set within a range including this portion, and pixels that satisfy this condition are set as verification targets.
The position where the visible light is irradiated is preferably closer to the food. As will be described later, since the food moves under the upper illumination device 14 while being placed on the conveyor 11, the food upper end portion and the lower end of the upper illumination device 14 are separated by a predetermined distance. It is necessary to keep

However, only the four elements of the light beam of the upper illumination device 14 cause the following problems. That is, if the food with irregularities on the surface is merely irradiated with visible light from the imaging camera 12 side, the shadow of the food may be recognized as a foreign object in the foreign object determination. For this reason, it is necessary to add a technique for eliminating the shadow caused by the unevenness of the food. The inventors of the present invention have found that the shadow cannot be completely erased by irradiating light from the imaging camera 12 with respect to the technique for eliminating the shadow. In order to solve this problem, in this embodiment, the lower illumination device 15 is provided as illumination located on the opposite side to the imaging camera 12.
As the lower illumination device 15, a device for irradiating infrared light was used. As will be described later, the food is placed on the conveyor 11 and transferred, the image is captured by the imaging camera 12 in the detection area A in the middle thereof, and the image is analyzed by the image processing device 13 to determine foreign matter detection. Therefore, the infrared light irradiated from the lower illumination device 15 is irradiated from the back surface of the food with the conveyor belt 11a on which the food is placed interposed therebetween.
At this time, the shadow based on the unevenness of the food disappears or becomes thin due to the structure of the conveyor belt 11a described above. For foods that are transparent or not transparent but transparent to infrared light, such as rice and udon, the shadow disappears completely. With regard to a material in which food hardly transmits infrared light, the infrared light reflected on the surrounding walls enters the unevenness of the food and erases the shadow.

The imaging camera 12 is a CCD camera or the like. The imaging camera 12 preferably increases the focal length of the lens so that the captured image can be focused even when food with large irregularities or the distance between the food surface and the bottom of the tray is large. Is easy. Preferably, when the visual field to be captured is 300 mm × 300 mm on the food surface and the above-mentioned distance is 40 mm at the maximum, good results are obtained when the focal length is in the range of 500 mm to 900 mm, preferably 500 mm to 800 mm.
In addition, as the number of pixels increases, one pixel can capture a smaller area. For example, the diameter of the hair is about 100 μm, and the insects have a size of 100 μm or more. Here, when viewing a 300 mm × 300 mm field of view with a 4 million pixel CCD camera, the pixel area is 150 μm × 150 μm. In this area, it is possible to capture hair or insects as foreign matter. Naturally, a 6 million pixel camera and a 12 million pixel camera have also been put into practical use and can be applied to this embodiment, but 4 million pixels are more preferable because of the disadvantage of being expensive.

The image processing device 13 is a device that detects foreign matter by performing image processing including binarization processing based on food images acquired by the imaging camera 12.
Here, as an example, a method for discriminating (1) black hair and (2) insect will be described.

(1) Black Hair Discrimination Method In the black-and-white binarization technique, one pixel that is stronger in black than the surroundings appears for all pixels, and this is used as a base point. Search for pixels with the same black anomaly around the base point. At this time, a threshold value is set in advance, and a pixel indicating a value exceeding the threshold value is defined as an abnormal pixel 41. The background pixel 40 is within the threshold. Then, the abnormal pixel 41 closest to the base point is recognized, and the next abnormal pixel 41 is searched with the pixel as a new reference pixel. This analysis is sequentially performed, and the case where the abnormal pixels 41 exist in a continuous linear shape as shown in FIG. 6A is recognized as a black fiber, that is, a fibrous foreign matter. Hair is also present in this black fiber. On the other hand, long foods such as hijiki are also included in the aforementioned black fiber material. Therefore, it is preferable to set in advance the fineness of the fibers recognized as foreign matters. As a result, long foods such as hijiki are excluded from foreign substances. In addition to hair, black fiber includes pig hair, black vegetable fiber, and the like. When only hair is detected, as shown in FIG. 6B, when the abnormal pixel 41 appears over a wide area, it is not determined as a foreign object.
On the other hand, the method for discriminating gray hair is the same as in the case of black described above, except that in the black-and-white binarization technique, for all pixels, one pixel that is whiter than the surroundings is started.

(2) Insect discriminating method In the black-and-white binarization technique, for every pixel, one pixel that is stronger in black than the surroundings appears, and this is used as a base point. Note that one pixel with strong black color is a part of an insect. Then, a pixel having the same black anomaly around the base point is searched. At this time, a threshold value is set in advance, and a pixel indicating a value exceeding the threshold value is defined as an abnormal pixel 41. The background pixel 40 is within the threshold. Then, the abnormal pixel 41 closest to the base point is recognized, and the next abnormal pixel 41 is searched with the pixel as a new reference pixel. This analysis is sequentially performed, and a case where abnormal pixels 41 exist so as to continuously form an area as shown in FIG. 6B is recognized as a black area foreign matter. Insects are also present in this black area foreign material. In addition to insects, black area foreign substances include burnt portions of food, black foreign substances other than insects, and the like. Note that, when detecting only foreign objects such as insects, as shown in FIG. 6A, when the abnormal pixels 41 appear in a continuous line shape, they are not determined as foreign objects in this embodiment.

Furthermore, in the present embodiment, since the foreign object detection device 10 is applied to food, a configuration unique to food is provided. Here, the case where the imaging camera 12 is installed in the upper part will be described as an example. At this time, the imaging camera 12, the upper illumination device 14, the transfer conveyor 11, and the lower illumination device 15 are arranged in this order from above. The bare food or the food stored in the tray is placed on the conveyor 11 and passes through the detection region A in a state of passing under the upper illumination device 14. A device for handling food needs to be kept clean from the viewpoint of hygiene. Therefore, it is necessary to clean at least a portion of the conveyor 11 where the food or tray comes into contact. Therefore, in this embodiment, as shown in FIG. 3, the cleaning means 30 for cleaning the surface of the conveyor belt 11a with water or a cleaning liquid is provided.
Here, when the upper illumination device 14 existing on the conveyor belt 11a and the belt have a space enough to allow food to pass therethrough, the upper illumination device 14 also has liquid when washed with water on the belt 11a. There is a risk that it will take. Accordingly, the upper illumination device 14 is configured to be movable, and is stored in a storage box 17 provided in the vicinity of the upper illumination device 14 at the time of cleaning, and the outer surface of the box is sealed so that water is not applied. The upper illumination device 14 is configured to move on the rail 18 (see FIG. 2). At this time, the storage box 17 may be the same as the box in which the imaging camera 12 is stored.

Further, when the cleaning means 30 is provided, there is a risk that the lower illuminating device 15 existing below the conveyor belt 11a will also get liquid. Therefore, a sealed storage box 19 that can store the lower illumination device 14 is provided. An infrared light transmission window 19a through which infrared light is transmitted is provided on the side of the storage box 19 that faces the belt surface. The infrared light transmission window 19a is preferably a transparent plastic plate, for example, a polycarbonate plate.
Further, in the storage box 19 isolated from the outside, heat is generated to always turn on infrared light. Therefore, a cooling means (not shown) is provided to cool the storage box 19. As the cooling means, for example, forced cooling means for introducing external air into the storage box and exhausting it to the outside by air piping can be cited.
As another embodiment, in the case of the configuration of the conveyor 11 shown in FIG. 4, an infrared light transmission plate 31 may be installed between the lower illumination device 15 and the conveyor 11. good. This prevents the cleaning liquid or water from being applied to the lighting device. Further, a cleaning device 32 that blows cleaning air toward the upper side surface of the infrared light transmitting plate 31 may be provided, whereby dust or the like falls from a gap between the transfer conveyors 11. In addition, since it is received by the infrared light transmitting plate 31 and blown off by the air from the cleaning device 32, it is possible to sufficiently maintain the light transmittance.

  Furthermore, the foreign object detection device 10 according to the present embodiment preferably includes a mechanism for excluding foreign-contaminated food from which foreign objects are detected. The food 20 is continuously supplied to the entrance of the conveyor 11 and conveyed. If it is determined from the result of foreign matter detection by the image processing device 13 that the foreign matter is mixed in the food, the food 20 The determination signal is transmitted to the food discharge mechanism installed on the exit side of the conveyor 11 so as to exclude only the foreign matter mixed food.

Next, as an example, a specific foreign object detection method using the foreign object detection device 10 described above will be shown.
The performance of the image processing device 13 included in the foreign object detection device 10 is set to such an extent that foreign matter attached to the food surface in the captured image can be detected. The image processing device 13 requires 0.8 to 1.5 seconds as the image processing time for detecting foreign matter on one screen, and can determine an area of 300 mm × 300 mm per minute from 75 to 40. Therefore, the speed of the conveyor 11 is 22.5 to 12 m / min. When one food item to be determined exists on one screen, 75 to 40 items can be determined. Since two trays can exist in a 300 mm × 300 mm area, 150 to 80 can be determined. This production amount is generally combined with the production amount for producing food and can be easily adopted in the conventional line.
According to this example, foreign substances such as hair (including black hair, white hair, and colored hair other than black), black / white pig hair other than hair, plant fibers, insects, and other foreign matters such as insects could be identified. Foods that are subject to foreign object detection include raw meat, processed meat, ham, sausage, meat dumplings, rice, udon, buckwheat, bread, sandwiches, lunch boxes, etc. , Vegetables, fruits, salads, fish fillets and squid rings.

Another embodiment of the conveyor 11 is shown in FIG. The conveyor 11 includes two conveyors 11A and 11B that are arranged close to each other in parallel, and the conveyors 11A and 11B are configured such that the transfer directions are opposite to each other. The two conveyors 11A and 11B are connected at the end on the same side, and are provided with a food reversing means 28 at the end. And the foodstuff 20 conveyed on the one conveyance conveyor 11A is reversed in the said inversion means 28, after the foreign material inspection is carried out in the detection area | region A, and the other side in the state which the back surface of the foodstuff 20 exposed When the sample is conveyed on the conveyor 11B and passes through the detection area A again, the foreign matter inspection of the back surface 20 is performed.
By adopting such a configuration, it becomes possible to detect foreign matter adhering to both the front and back surfaces of the food 20 with a single detection means, and the inspection time can be shortened.

  The present invention minimizes misrecognition of foreign matters by providing illumination means for irradiating illumination light from at least two directions across the food, even for food having a complicated shape, and binarization processing By detecting the continuity in the abnormal pixels of the binary image obtained in this way, it is possible to detect foreign matters such as hair and insects that have been difficult to detect with high accuracy, and to detect various foreign matters mixed in any food. It is possible to detect.

1 is a front view of a foreign object detection device according to an embodiment of the present invention. It is a side view of the foreign material detection apparatus shown in FIG. It is a schematic front view which shows the foreign material detection apparatus provided with the washing | cleaning means. It is a schematic front view which shows the foreign material detection apparatus which concerns on another Example. It is a schematic plan view which shows the other Example of a conveyance conveyor. It is a figure which shows the binary image obtained with the image processing apparatus of a present Example, and is a figure which shows a fibrous foreign material. It is a figure which shows the binary image obtained with the image processing apparatus of a present Example, and is a figure which shows an area foreign material.

Claims (10)

In the foreign matter detection device that detects foreign matter mixed in food,
Conveying means for conveying the food to the foreign substance detection area, imaging means for capturing a front image of the food from one side of the detection area, and the food from one or both sides of the inspection area across the conveying means Illuminating means for irradiating illumination light, and image processing means for performing image processing including binarization processing on the captured table image,
The image processing means binarizes the captured table image, and after detecting one abnormal pixel exceeding a preset threshold value in the obtained binary image, an abnormality of a pixel adjacent to the one pixel is detected. If abnormal pixels are detected, the detection of abnormalities in the pixels adjacent to the detected abnormal pixels is repeated repeatedly, and abnormal pixels appear continuously based on the one pixel. A foreign object detection device for food, wherein the foreign object detection unit determines that the object is a foreign object.   The food foreign matter detection device according to claim 1, wherein the abnormal pixel is determined as a fibrous foreign matter such as hair when the abnormal pixel appears in a continuous line shape based on the one pixel.   The food foreign object detection device according to claim 1, wherein the abnormal pixel is determined to be an area foreign object when it appears in a continuous area shape based on the one pixel.   Of the illumination means, the first illumination means located on the imaging means side is a means for irradiating visible light, and the second illumination means located on the other side is a means for irradiating infrared light. The foreign object detection device for food according to claim 1, wherein:   5. The foreign object detection device for food according to claim 4, wherein the first illumination unit appropriately selects one of red, blue, green, and white light as a mixed color according to the type of the food. .   2. The foreign object detection apparatus for food according to claim 1, wherein the imaging means is a CCD camera, and a focal length of a lens of the CCD camera is in a range of 500 mm to 900 mm.   The transport unit includes a cleaning unit that cleans a portion that comes into contact with the food, and at least one of the illumination units is configured to be movable, and the illumination unit is separated from the transport unit during cleaning by the cleaning unit. The food foreign matter detection device according to claim 1, wherein the food foreign matter detection device is stored in a storage box provided.   The second illumination unit is housed in a sealed storage box formed of a material that transmits infrared light on the side facing the transport unit, and the sealed storage box includes a cooling unit that cools the inside. The foreign object detection device for food according to claim 4.   The conveying means has two conveyor belts arranged in parallel, and the conveying directions of the two conveyor belts are opposite to each other, and the food reversing means is disposed at the end of the conveyor belt. The foreign matter on the front surface of the food is detected on one conveyor, and the foreign matter on the back surface of the food is detected on another conveyor. Detection device.   2. The food foreign matter detection apparatus according to claim 1, wherein the foreign matter is a substance mainly composed of protein.

Images