Foreign matter removal system and foreign matter removal method

Abstract

To provide a system for removing foreign matter from foods including multiple kinds of foreign matter other than normal foods.SOLUTION: Firstly, by using a discrimination device 101 located above a conveyance surface of a conveyance device 110, a foreign matter removal system 100 discriminates, on the basis of a discrimination model generated by machine learning, foreign matter contained in foods conveyed by the conveyance device 110 (S301). Secondly, the foreign matter removal system 100 blows air to the discriminated foreign matter from air nozzles 102 connected to the discrimination device 101 on the conveyance surface of the conveyance device 110 and discharges the discriminated foreign matter from the foods (S302). Thirdly, the foreign matter removal system 100 stores the discharged foreign matter in a container 103 located near a terminal end of the conveyance device 110 (S303).SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a foreign object removal device and a foreign object removal method, and more particularly to a device and a method for removing foreign objects contained in food. [Background technology]

[0002] Various foreign objects are found on landed seafood, and these are removed during the distribution process before being sold to consumers. In addition to unintended seafood, foreign objects include pebbles, plastic, scrap iron, paint, and tobacco. As such a wide variety of foreign objects are expected, it is difficult to automate the removal of foreign objects, and in many cases, they are still visually inspected and removed manually by humans. For example, multiple workers are stationed on a line equipped with a conveying device to remove foreign objects from the seafood being transported.

[0003] Conventionally, there is a method for sorting and removing foreign objects by passing food materials falling from the end of a conveyor belt in front of a comparison color plate, detecting foreign objects based on the color difference between the food materials and the comparison color plate, and spraying air at the detected foreign objects to change their falling trajectory.

[0004] Patent Document 1 discloses a method applicable to agricultural products such as nuts and beans to address the problem of inconsistent trajectories of raw materials falling from the end of a conveyor belt, in which the falling trajectory is stabilized by increasing the conveying speed of the conveyor belt and making the rollers at the end of the conveyor belt smaller in diameter than the rollers at the start of the conveyor belt. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 8-108146 Summary of the Invention [Problem to be solved by the invention]

[0006] However, many foods cannot be considered granular like marine products. In addition, even granular foods, whether marine or agricultural, often contain multiple types of foreign objects, and in such cases, the trajectory of the foreign objects varies depending on the object, so problems remain with conventional methods.

[0007] The present invention has been made in view of the above-mentioned points, and an object of the present invention is to provide an apparatus or method for removing a plurality of types of foreign matter from food that contains the foreign matter in addition to normal food. Here, in this specification, "food" includes not only fresh food but also processed food. [Means for solving the problem]

[0008] In order to achieve this objective, a first aspect of the present invention is a foreign object removal device comprising a discrimination device arranged above a conveying surface of a conveying device for discriminating foreign objects contained in food conveyed by the conveying device based on a discrimination model generated by machine learning, a plurality of air nozzles for injecting air onto the conveying surface at the discriminated foreign objects to expel the discriminated foreign objects from the food, and a storage container arranged near the end of the conveying device for storing the expelled foreign objects.

[0009] A second aspect of the present invention is the foreign matter removal device of the first aspect, wherein the plurality of air nozzles are arranged to have an angle of 10° or more and less than 30° with respect to a conveying surface of the conveying device.

[0010] A third aspect of the present invention is the foreign matter removal device according to the first aspect, wherein the discrimination device has an image capture device for capturing an image of the transport surface from above.

[0011] A fourth aspect of the present invention is the foreign matter removal device according to any one of the first to third aspects, wherein compressed air is supplied to the plurality of air nozzles.

[0012] A fifth aspect of the present invention is the foreign matter removal device of any one of the first to fourth aspects, wherein the storage container can be placed above a subsequent transport device that is placed downstream of the transport device.

[0013] In addition, a sixth aspect of the present invention is a foreign matter removal method including the steps of using a discrimination device arranged above a conveying surface of a conveying device to discriminate foreign matter contained in food conveyed by the conveying device based on a discrimination model generated by machine learning, spraying air from an air nozzle connected to the discrimination device on the conveying surface toward the discriminated foreign matter to expel the discriminated foreign matter from the food, and storing the expelled foreign matter in a storage container arranged near the end of the conveying device.

[0014] A seventh aspect of the present invention is the foreign matter removal method according to the sixth aspect, wherein a conveying speed of the conveying device is equal to or greater than 5 m / min and less than 17 m / min or less.

[0015] An eighth aspect of the present invention is the foreign matter removing method according to the seventh aspect, wherein the conveying speed is equal to or greater than 7 m / min and less than 10 m / min or less.

[0016] A ninth aspect of the present invention is a foreign matter removal method according to any one of the sixth to eighth aspects, wherein the storage container is positioned above a subsequent conveying device that is arranged downstream of the conveying device.

[0017] In addition, a tenth aspect of the present invention is a method for processing or manufacturing food, comprising the steps of using a discrimination device arranged above a conveying surface of a conveying device to discriminate foreign matter contained in food conveyed by the conveying device based on a discrimination model generated by machine learning, spraying air from an air nozzle connected to the discrimination device on the conveying surface toward the discriminated foreign matter to expel the discriminated foreign matter from the food, and storing the expelled foreign matter in a storage container arranged near the end of the conveying device. Effect of the Invention

[0018] According to one aspect of the present invention, unlike conventional methods, rather than spraying air at foreign objects falling from the end of the conveying device, air is sprayed onto the conveying surface of the conveying device at identified foreign objects to remove them from the food. This eliminates problems caused by different falling trajectories depending on the type of foreign object, and allows for stable removal even if multiple different types of foreign objects are mixed into the food. [Brief description of the drawings]

[0019] [Figure 1] 1 is a diagram showing a foreign matter removal device according to an embodiment of the present invention; [Diagram 2] FIG. 2 is a top view of the foreign matter removal device of FIG. [Diagram 3] 1 is a diagram showing a general flow of a foreign matter removal method according to an embodiment of the present invention. [Figure 4] 1 is a diagram showing a specific example of the flow of a foreign matter removal method according to an embodiment of the present invention. [Diagram 5] 1 is a diagram showing a specific example of the flow of a foreign matter removal method according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

[0021] A foreign matter removal apparatus according to one embodiment of the present invention is shown in Figure 1. Foreign matter removal apparatus 100 includes a discrimination device 101 for discriminating foreign matter contained in food conveyed by conveying device 110 based on a discrimination model generated by machine learning, a plurality of air nozzles 102 for ejecting foreign matter discriminated by discrimination device 101 from the food by injecting air onto the conveying surface of conveying device 110, and a storage container 103 for storing the ejected foreign matter.

[0022] In the example of FIG. 1, the discrimination device 101 is disposed above the conveying surface of the conveying device 110, which is a conveying belt, and therefore can be installed without requiring additional space in the conveying direction of the conveying device 110. In one example, the discrimination device 101 may have at least a photographing device for photographing the food conveyed by the conveying device 110 from above, and a processor for discriminating foreign objects based on a discrimination model generated by machine learning using the photographed image as an input. The processor of the discrimination device 101 may further perform control for operating the multiple air nozzles 102, but such control may be performed by a separate processor. In addition, a server (not shown) accessible by the discrimination device 101 via an IP network such as the Internet may perform discrimination processing based on a discrimination model, and the discrimination result may be transmitted from the server to the discrimination device 101, and the air nozzle 102 may be controlled according to the discrimination result. In this case, the discrimination device 101 has a communication interface for transmitting and receiving data to and from the server, in addition to the photographing device. In any case, the discrimination device 101 may be a device for discriminating foreign matter contained in food based on a discrimination model generated by machine learning, and some of the processes required for the discrimination process may be performed by other devices. The layout relationship of the components of the discrimination device 101 may be appropriately determined, so details will not be described.

[0023] The multiple air nozzles 102 can be arranged to form an angle of 10° or more and less than 30° with respect to the conveying surface of the conveying device 110, and are preferably arranged to form an angle of 15° or more and less than 25°.

[0024] The storage container 103 is disposed near the end of the conveying device 110. When a subsequent conveying device 120 having a conveying surface at a starting end lower than the conveying surface at the end of the conveying device 110 is used downstream of the conveying device 110, it is preferable to dispose the storage container 103 near the end of the conveying device 110 and above the subsequent conveying device 120. In this manner, the storage container 103 as well as the discrimination device 101 do not require additional space in the conveying direction, and can be installed in a wide variety of factories, which promotes labor saving in foreign object removal work.

[0025] Fig. 2 is a top view of the foreign matter removal apparatus of Fig. 1. For simplicity, wiring for operating the multiple air nozzles 102 and the subsequent transport device 120 are omitted.

[0026] FIG. 3 shows the flow of a foreign matter removal method according to one embodiment of the present invention. First, the foreign matter removal device 100 uses the discrimination device 101 arranged above the conveying surface of the conveying device 110 to discriminate foreign matter contained in food conveyed by the conveying device 110 based on a discrimination model generated by machine learning (S301). The discrimination model can be generated by machine learning, for example, by supervised learning. The generated discrimination model may be one that has only learned whether an item is normal or not, or one that has learned normal items and multiple types of foreign matter as their respective classes. In the former case, objects other than those discriminated as normal items can be discriminated as foreign matter. Multiple types of foreign matter can also be learned together as one class.

[0027] Next, foreign matter removal apparatus 100 injects air from air nozzle 102 connected to discrimination device 101 onto the conveying surface of conveying device 110 toward the identified foreign matter, thereby removing the identified foreign matter from the food (S302). Compressed air can be supplied to the air nozzle to inject blower air. One example of a method for supplying compressed air to the air nozzle is to connect a solenoid valve to each of a plurality of nozzles, and connect the solenoid valves to a manifold that distributes compressed air supplied from a compressor.

[0028] Then, the foreign matter removal device 100 stores the discharged foreign matter in a storage container 103 arranged near the end of the transport device 110 (S303).

[0029] Unlike conventional methods, foreign object removal device 100 does not spray air at falling foreign objects, so there is no problem caused by different falling trajectories depending on the type of foreign object. Therefore, even if different types of foreign objects are mixed into food, they can be stably removed by air nozzle 102.

[0030] The storage container 103 can be disposed above the subsequent conveying device 120 disposed downstream of the conveying device 110. In this case, no additional space is required in the conveying direction, and the number of workers required for the conveying device 110 can be reduced, while workers can continue to remove foreign objects and perform other tasks at the downstream subsequent conveying device 120, thereby achieving food inspection that coexists with humans.

[0031] In addition, while in conventional methods the conveying speed had to be relatively high to make the foreign objects fall along the desired trajectory, the foreign object removal device 100 does not require this, and the conveying speeds of the conveying device 110 and the subsequent conveying device 120 can be set to speeds at which workers can work, making it possible to practically implement a line production method that coexists with humans. Specifically, the conveying speed of the conveying device 110 can be set to 5 m / min or more and less than 17 m / min or less, and can be set to 7 m / min or more and less than 10 m / min or less. The conveying speed of the subsequent conveying device 120 can also be set to a similar speed.

[0032] Furthermore, since the foreign object removal device 100 does not spray air at falling foreign objects, it is highly effective in removing foreign objects from wet foods that do not follow a consistent trajectory. In this specification, "wet foods" refers to foods with a moisture content of over 50%. For example, dried whitebait generally has a moisture content of about 70%, and the foreign object removal device 100 can be applied to removing foreign objects from foods with a moisture content of 70% or more. Furthermore, fibrous foreign objects such as hairs also do not follow a consistent falling trajectory, so the foreign object removal device 100 is more effective than conventional methods in removing such foreign objects.

[0033] In addition, in the foreign matter removal device 100, a machine learning model is used to directly or indirectly discriminate foreign matter, and if the conveying speed of the conveying device 110 is high, this adversely affects the discrimination accuracy of the machine learning model. Therefore, in the conventional method that requires a high conveying speed to allow the foreign matter to draw a desired trajectory, it is not appropriate to adopt machine learning. As described above, spraying air onto the foreign matter on the conveying surface of the conveying device 110 makes it possible to discharge multiple types of foreign matter with high accuracy and to reduce the conveying speed. In addition, when the conveying speed is reduced, this leads to improved discrimination accuracy of the machine learning model. For example, if the discrimination accuracy of the machine learning model is improved, it becomes possible to discriminate the presence of an object that is the same color as the color of the conveying surface. In the conventional method, a normal product or a foreign matter is discriminated based on the color difference with the comparison color plate in the background, so that it is difficult to remove foreign matter of the same color, whereas in the method according to one embodiment of the present invention, such foreign matter can be removed by reducing the conveying speed of the conveying device 110 to 5 m / min or more and less than 17 m / min or less as described above.

[0034] 4 and 5 show a specific example of a foreign matter removal method according to an embodiment of the present invention. First, a camera included in the discrimination device 101 captures an image of the food from above the conveying surface of the conveying device 110 (S401). At this time, the quality of the captured image can be improved and discrimination accuracy can be improved by covering the camera with the housing of the discrimination device 110 except for the side surface facing the conveying surface. When generating a discrimination model by machine learning, it is also preferable to use images captured in a similar shooting environment as training data.

[0035] Next, the discrimination device 101 detects a foreign object using a machine learning model with the acquired image as an input (S402). If no foreign object is detected, the discrimination device 101 takes a new image of the conveying surface of the conveying device 110 and repeats the same process (S403). At this time, the image capturing by the camera may be performed continuously or intermittently without waiting for the result of foreign object detection.

[0036] When a foreign object is detected (S403), the discrimination device 101 calculates a two-dimensional coordinate with the transport direction of the foreign object as the x-axis and the direction perpendicular thereto as the y-axis (S501). According to the value of the y-coordinate, the nozzle number of the air nozzle to be operated is selected from the multiple air nozzles 102 arranged in the y-direction (S502). In FIG. 2, 18 air nozzles are arranged in the y-direction, and an air nozzle is further arranged at each end at an angle of 40° inward. For example, if the y-coordinate of the foreign object is between two air nozzles, two adjacent air nozzles arranged at positions close to the y-coordinate of the foreign object may be selected as the operation targets. In addition, for a large foreign object, multiple air nozzles may be selected as the operation targets to increase the amount of air blown onto the foreign object.

[0037] By increasing the number of air nozzles, foreign objects can be targeted with high precision, and the loss rate of normal products can be reduced. For this purpose, it is preferable that the dimensions of the air nozzle nozzle outlet are a small diameter of φ0.5 or more, or less than φ0.8 or less. If the outlet is not circular, it is preferable that the length of the long side is 0.5 mm or more, or less than 0.8 mm or less.

[0038] Then, a waiting time until the selected air nozzle or nozzles are operated is calculated according to the x coordinate of the foreign object (S503), and when the waiting time has elapsed (S504), air is sprayed (S505).

[0039] In the above description, with reference to Figures 3 to 5, embodiments of the present invention have been described as a foreign matter removal method, but these embodiments may also be applied to new food processing or manufacturing methods that involve removing foreign matter.

[0040] In the above-mentioned embodiments, it should be noted that unless there is a statement of "only" such as "based only on", "depending only on", or "only in the case of", it is assumed in this specification that additional information may be taken into consideration. Also, as an example, it should be noted that the statement "do b when a" does not necessarily mean "always do b when a" or "do b immediately after a" unless expressly stated. Also, the statement "each a constituting A" does not necessarily mean that A is composed of multiple components, but includes the case where the component is singular.

[0041] Also, just to be clear, even if there is an aspect of a method, program, terminal, device, server or system (hereinafter referred to as a "method, etc.") that performs an operation different from that described in this specification, each aspect of the present invention is directed to an operation that is the same as any of the operations described in this specification, and the existence of an operation different from that described in this specification does not make the method, etc. outside the scope of each aspect of the present invention.

[0042] Furthermore, the "start" and "end" shown in Figures 3 to 5 are merely examples, and do not mean that the method of this embodiment necessarily starts or ends in the steps shown. [Explanation of symbols]

[0043] 100 Foreign matter removal device 101 Discrimination device 102 Air nozzle 103 Containment Container 110 Conveyor 120 Subsequent conveying device

Claims

[Claim 1] a discrimination device disposed above a conveying surface of the conveying device for discriminating foreign matter contained in food conveyed by the conveying device based on a discrimination model generated by machine learning; a plurality of air nozzles for ejecting air onto the conveying surface against the identified foreign matter to remove the identified foreign matter from the food; a container disposed near the end of the conveying device for accommodating the discharged foreign matter; A foreign body removal device comprising:

Images