Special material recognizer and removal confirmation device
By using machine learning models and multispectral imaging technology to identify foreign materials, the false positive problem in foreign material identification equipment has been solved, ensuring reliable identification and removal of foreign materials and improving product quality and equipment safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2026-03-27
AI Technical Summary
Existing foreign material identification devices have too many false positive signals and it is difficult to confirm whether foreign materials have been truly removed from the product flow, which affects product quality and equipment safety.
Machine learning models are used to analyze images of foreign materials, and multispectral or hyperspectral imaging technology is combined to identify and classify foreign materials, generate reliable alarm signals, and confirm the removal status of foreign materials through a camera system.
Reduce false positive alarms, ensure reliable identification and removal of foreign materials, improve product quality and equipment safety, and reduce the inconvenience and equipment damage caused by false alarms.
Smart Images
Figure CN121753078A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present invention relates to product sorting, and in particular to the identification and removal of foreign material during product sorting. BACKGROUND
[0002] In the field of food sorting and other bulk product sorting, the removal of foreign material is often a primary concern. Foreign material can cause damage to the product and equipment, and even more foreign material in the product as it continues to travel through the process line. Therefore, the removal of these foreign objects is imperative. Process line operators and production companies consider this so important that they want a specific alarm from the sorting equipment when foreign material is identified, but these alarms can become a nuisance when there are too many false alarms or false positive signals. Depending on the product being processed and the processing conditions, the number of false positives can exceed the true foreign material conditions by a ratio of hundreds to one. A device is required to verify and qualify the signal indicating foreign material to ensure that an alarm is issued for true foreign material, while other signals should be ignored. This device that secondarily verifies the signal can take longer to analyze the signal in more detail and use other more time consuming analysis techniques, which is not possible in a sorting machine that must make multiple decisions as the product travels to the rejection point.
[0003] A second related problem is verifying that the foreign material has actually been removed from the product stream. This creates the following problem in bulk sorting: the product is randomly placed on the infeed transport system, and after flowing through the sorting / rejection zone, the position of any item relative to the flow and other products will have changed. If the object is rejected from the flow, its position in the reject flow will be randomized, as its appearance on the rejected product transport device can be immediate, or can be delayed, depending on where it randomly landed on the transport device and the speed of the transport system. The orientation of the object will not match the orientation when it was inspected, and as a three-dimensional object, the visible aspects of the object can be very different from the aspects that were visible when initially inspected. SUMMARY
[0004] According to the invention, there is provided a method for identifying at least one image of foreign material from a plurality of images of suspected foreign material from a sorting system, the method comprising: obtaining at least one image of suspected foreign material detected in a stream of good product containing a quantity of foreign material; performing image analysis on the at least one image of suspected foreign material; and based on the image analysis, classifying the suspected foreign material as true foreign material or non-foreign material.
[0005] The method can further comprise generating a signal indicating that the suspected foreign material has been classified as true foreign material when the suspected foreign material is classified as true foreign material.
[0006] The method may further include: generating a signal indicating that the suspected foreign material has been classified as a non-foreign material when the suspected foreign material is classified as a non-foreign material.
[0007] Performing image analysis can include using machine learning models trained on images of both real and non-alien materials.
[0008] Performing image analysis may include performing at least one of the following: image processing functions for shape recognition, image processing functions for size recognition, image processing functions for color recognition, image processing functions for texture recognition, and image processing functions for classifying materials using multivariate image analysis.
[0009] The at least one image of the heterogeneous material can be a multispectral or hyperspectral image with spectral bands in the range of 350 nm to 1700 nm.
[0010] The method may also include storing images of any suspected foreign material that is classified as a true foreign material.
[0011] The method may also include actuating a secondary rejection device using a signal indicating that a suspected foreign material has been classified as a genuine foreign material.
[0012] The method may also include providing at least one operator with an image of any suspected foreign material that is classified as a genuine foreign material.
[0013] According to the present invention, an apparatus is also provided for identifying at least one image of a foreign material from a plurality of images of suspected foreign materials from a sorting system, the apparatus comprising: components for obtaining at least one image of a suspected foreign material detected in a stream of good products containing a certain number of foreign materials; components for performing image analysis on the at least one image of the suspected foreign material; and components for classifying the suspected foreign material as a true foreign material or a non-foreign material based on the image analysis.
[0014] The device may also include a component for generating a signal indicating that a suspected foreign material has been classified as a genuine foreign material when the suspected foreign material is classified as a genuine foreign material.
[0015] The device may also include a component for generating a signal indicating that a suspected foreign material has been classified as a non-foreign material when the suspected foreign material is classified as a non-foreign material.
[0016] The component for performing image analysis may include components for using a machine learning model trained on images of both real and non-alien materials.
[0017] The components for performing image analysis may include components for performing at least one of the following: a shape recognition processing function, a size recognition image processing function, a color recognition image processing function, a texture recognition image processing function, and a material classification function using multivariate image analysis.
[0018] The at least one image of the heterogeneous material can be a multispectral or hyperspectral image with spectral bands in the range of 350 nm to 1700 nm.
[0019] The device may also include a component for storing images of any suspected foreign material that is classified as a true foreign material.
[0020] The device may also include components for actuating a secondary rejection device using a signal indicating that a suspected foreign material has been classified as a genuine foreign material.
[0021] The device may also include a component for providing at least one operator with an image of any suspected foreign material that is classified as a genuine foreign material.
[0022] According to the present invention, a method for confirming the removal of foreign matter from a product stream during a sorting process is also provided, the method comprising: obtaining at least one image of at least one product removed from the product stream by a sorting system; comparing at least one image of the at least one removed product with at least one image of a foreign matter in the product stream identified by the sorting system for removal; and determining that the foreign matter has been successfully removed from the product stream when the at least one image of the foreign matter matches the at least one image of the at least one product removed from the product stream by the sorting system. The at least one image of the foreign matter identified by the sorting system for removal in the product stream can be obtained from the sorting system. The at least one image of the at least one product removed from the product stream by the sorting system can be obtained using a camera.
[0023] Obtaining at least one image of at least one product removed from the product stream by the sorting system may include recording the at least one image of the at least one product using a camera positioned along the rejection path leaving the sorting system.
[0024] Obtaining at least one image of at least one product removed from the product stream by the sorting system may include obtaining multiple images of multiple objects removed from the product stream.
[0025] The method may also include signaling confirmation of successful removal.
[0026] The method may also include obtaining at least one image of at least one product retained in the product stream by the sorting system.
[0027] The method may further include: comparing at least one image of at least one retained product with at least one image of a foreign object identified by a sorting system for removal in the product stream; and determining that the foreign object was not successfully removed from the product stream when the at least one image of the foreign object matches at least one image of at least one product retained in the product stream by the sorting system.
[0028] The method may also include signaling confirmation of unsuccessful removal.
[0029] Obtaining at least one image of at least one product retained in the product flow by the sorting system may include recording the at least one image of the at least one product using a camera positioned along the acceptance path leaving the sorting system.
[0030] According to the present invention, a method is provided for confirming the removal of foreign matter from a product stream during a sorting process, the method comprising: obtaining at least one image of at least one product removed from the product stream by a sorting system; comparing the at least one image of the at least one removed product with at least one image of a foreign matter in the product stream identified by the sorting system for removal; and determining that the foreign matter was not successfully removed from the product stream when the at least one image of the foreign matter does not match the at least one image of the at least one product removed from the product stream by the sorting system.
[0031] Determining that a foreign object was not successfully removed from the product stream can include determining that the foreign object was not successfully removed from the product stream at the end of a time period.
[0032] The time period can begin when the sorting system identifies a foreign object in the product stream. This time period can depend on the flow rate of the products through the sorting system and / or be based on the estimated time it has taken for the products to have passed through the sorting system.
[0033] Any method steps and apparatus features described herein may be used alone or in combination in any order according to any aspect of the invention.
[0034] According to the present invention, an apparatus is provided for confirming the removal of foreign matter from a product stream during a sorting process. The apparatus includes: components for obtaining at least one image of at least one product removed from the product stream by a sorting system; components for comparing the at least one image of the at least one removed product with at least one image of a foreign matter in the product stream identified by the sorting system for removal; and components for determining that the foreign matter has been successfully removed from the product stream when the at least one image of the foreign matter matches the at least one image of the at least one product removed from the product stream by the sorting system.
[0035] The components for obtaining at least one image of at least one product removed from the product stream by the sorting system may include components for recording said at least one image of at least one product using a camera positioned along the rejection path leaving the sorting system.
[0036] The component for obtaining at least one image of at least one product removed from the product stream by the sorting system may include a component for obtaining multiple images of multiple objects removed from the product stream.
[0037] The device may also include a component for signaling confirmation of successful removal.
[0038] The apparatus may also include a component for obtaining at least one image of at least one product retained in the product stream by the sorting system.
[0039] The apparatus may further include: a component for comparing the at least one image of at least one retained product with at least one image of a foreign object identified by the sorting system for removal in the product stream; and a component for determining that the foreign object has not been successfully removed from the product stream when the at least one image of the foreign object matches the at least one image of at least one product retained in the product stream by the sorting system.
[0040] The device may also include a component for signaling a determination that was not successfully removed.
[0041] The components for obtaining at least one image of at least one product retained in the product flow by the sorting system may include components for recording said at least one image of at least one product using a camera positioned along the acceptance path leaving the sorting system.
[0042] According to the present invention, an apparatus for confirming the removal of foreign matter from a product stream during a sorting process is also provided, the apparatus comprising: components for obtaining at least one image of at least one product removed from the product stream by a sorting system; components for comparing the at least one image of the at least one removed product with at least one image of a foreign matter in the product stream identified by the sorting system for removal; and components for determining that the foreign matter has not been successfully removed from the product stream when the at least one image of the foreign matter does not match the at least one image of the at least one product removed from the product stream by the sorting system.
[0043] Components used to determine that foreign matter has not been successfully removed from the product stream may include components used to determine at the end of a time period that foreign matter has not been successfully removed from the product stream.
[0044] This time period can begin when the sorting system identifies foreign objects in the product flow.
[0045] According to the present invention, a computer-readable medium having instructions thereon is provided, which, when executed by a processor, cause the processor to perform any of the methods mentioned above.
[0046] Given the reliable signal from the post-sorting qualifier, a secondary, similar qualifier checks the rejected product stream to match the sorted signal with the signal of the item transported from the rejection system. If a match is detected, the sorting system can definitively confirm the rejection of the foreign material. A similar secondary qualifier system can check the products on the receiving side of the sorting machine's rejection system to prevent foreign materials from being missed and can issue an alarm indicating the presence of foreign materials in acceptable products. If no foreign material is identified in either stream, it is possible that the foreign material has been rejected and escaped from the system in some way (bounce / bounce), or has been damaged beyond recognition by the rejection equipment, or has been obscured by other products in either the rejection or receiving stream.
[0047] For a batch of products passing through an inspection area, inspection methods / equipment can be used to detect foreign materials present or present in the batch of products passing through the inspection area. When a foreign material is detected with high certainty, a rejection method / equipment can be activated via a signal from the inspection method / equipment, which rejects the detected material onto a transport vehicle leading to a second inspection method / equipment. The second inspection equipment can detect the rejected object and match it with the object detected by the first inspection method / equipment. A method can also be provided to communicate that a foreign material has been rejected and to confirm that it was detected by the second inspection method / equipment.
[0048] An optional third inspection method / equipment can detect the batch product flow after the first inspection and rejection method / equipment and attempt to match any detected foreign materials in case they are missed or the rejection method / equipment fails to remove them from the batch flow.
[0049] The testing method can be a camera system, which can be a monochrome, color, color camera with an infrared channel, hyperspectral camera, and / or multispectral camera. The illumination used for the testing system can be monochrome, broadband, or multiple narrowband, etc. The illumination source can be constant or pulsed.
[0050] Inspection methods can refer to the use of sensors that reflect or transmit electromagnetic radiation (light) from the product being inspected. The reflected or transmitted radiation is focused on the sensor to be converted into a signal to be digitized and to form a digital image of the product being inspected.
[0051] Inspection methods can combine multiple views of the product flow to inspect the top and bottom surfaces of the product. Inspection can include multiple views of the product flow from different angles to inspect multiple aspects of the product.
[0052] Depending on the material and product being inspected, the electromagnetic (EM) radiation or light used to illuminate the product for a camera or detector can be in the UV, visible, and infrared regions of the spectrum, ranging from 380 nm to 1700 nm. In cases where inspection is performed by transmitting EM radiation through the product, the spectrum extends into the X-ray band and can include dual-energy X-ray imaging techniques.
[0053] Foreign material inspection can be achieved through various techniques using image processing and analysis of images output from inspection sensors. One or more second inspection units and an optional third inspection unit can be adequately matched with the sensing technology on the main sensor so that images from subsequent inspection equipment can be meaningfully compared with the main inspection system.
[0054] It is possible to classify objects in an image using a neural network classifier trained on images from a master inspection device that has sufficient spectral channels and resolution to identify dissimilar materials relative to good products and other products present on the production line.
[0055] A cloud-based repository of image data for creating the classifier model is best suited for this invention because images can be collected from multiple connected locations across different geographic locations, and the more species used in model training, the more robust the classifier becomes. Similarly, the classifier model learns from images of good products and products that are not heterogeneous, thereby reducing the chance of overlearning and bias towards specific species or local conditions by collecting images of heterogeneous products from a broad geographic base and from many testing machines.
[0056] According to one aspect of the invention, an apparatus for identifying at least one image of a foreign material from a plurality of images of suspected foreign materials from a sorting system is also provided, comprising: a sorting system for detecting foreign materials from a stream of good products containing a number of foreign materials; a component for transmitting an image of the detected foreign material; and a computer having a memory and processing capabilities suitable for image analysis and a communication component suitable for receiving the transmitted image.
[0057] The computer also includes communications for notifying the true or false identification of the foreign material based on the image analysis signal.
[0058] The image analysis process may include an inference engine and a machine learning model trained on images of both real and non-alien materials.
[0059] Image analysis processes can include image processing functions for recognizing shapes, recognizing dimensions, recognizing colors, recognizing textures, and classifying materials using multivariate image analysis.
[0060] The transmitted images of the heterogeneous material can be multispectral or hyperspectral images with spectral bands in the range of 350 nm to 1700 nm.
[0061] The truly recognized images can be transmitted to a database and stored.
[0062] The communication for true identification may include signals that actuate the secondary rejection device.
[0063] The image, which indicates true recognition, can be communicated to the operator.
[0064] The computing power and processing capabilities may include at least one GPU graphics processing unit or APU accelerated processing unit for implementing the neural network associated with the machine learning model.
[0065] GPUs or APUs can be integrated into a central processing unit (CPU).
[0066] According to one aspect of the invention, an apparatus for confirming the output of a sorting system (11) is also provided, comprising a component (12) for acquiring an image of a rejected object from the sorting system, a communication component (13) for acquiring an image of an accepted object from the sorting system, at least one camera and lighting device (14) for recording an image of a product rejected by the sorting system, optionally including at least one camera and lighting device (15) for recording an image of a product accepted by the sorting system, an image processing computer (16) for matching an image of an object intended to be rejected with an image of an object recorded by the camera (14) and confirming that it is not found in the image recorded by the camera (15), and a communication method for signaling the success or failure of the sorting rejection. Attached Figure Description
[0067] Embodiments of the invention will be described by way of example only with reference to the accompanying drawings, wherein:
[0068] Figure 1 This is a representation of an apparatus according to an embodiment of the present invention for identifying at least one image of a foreign material from a plurality of images of suspected foreign materials from a sorting system;
[0069] Figure 2 This is a representation of an apparatus according to an embodiment of the present invention for identifying at least one image of a foreign material from a plurality of images of suspected foreign materials from a sorting system;
[0070] Figure 3 This is a representation of an apparatus according to an embodiment of the present invention for confirming the removal of foreign matter from the product stream during a sorting process, and / or for identifying at least one image of a foreign matter from a plurality of images of suspected foreign matter from a sorting system;
[0071] Figure 4This is a representation of an apparatus according to an embodiment of the present invention for confirming the removal of foreign matter from the product stream during a sorting process, and / or for identifying at least one image of a foreign matter from a plurality of images of suspected foreign matter from a sorting system;
[0072] Figure 5 This is a representation of an apparatus according to an embodiment of the present invention for confirming the removal of foreign matter from the product stream during a sorting process, and / or for identifying at least one image of a foreign matter from a plurality of images of suspected foreign matter from a sorting system. Detailed Implementation
[0073] The identification of foreign materials (also referred to herein as foreign bodies or foreign objects) has been a requirement in the sorting and processing industry for many years, but until recently, the occurrence of too many false positives has made this signal more of a nuisance than a valuable function. One aspect of this invention involves using neural network inference techniques for pattern recognition and image processing to learn and confidently distinguish false signals from actual foreign body signals. Positive signals and the same neural network techniques can then be used to perform image analysis on the rejected and accepted product streams to confidently confirm the rejection of foreign bodies, and if this cannot be confirmed, then a warning is issued. This invention can also be extended to compiling statistics from foreign materials and generating reports on product quality and traceability reports on what has been removed from the product, etc.
[0074] This invention provides an apparatus for definitively identifying foreign materials from normal good products and other signals that cause false positives, allowing for a reliable and robust foreign material alarm signal to the processing plant and external systems. This signal can then be used to trigger other inspection and limiting systems to locate and identify objects identical to those rejected in the rejection transmission system at a later time, definitively confirming the capture of the foreign material in the rejection system. An optional secondary limiter can be used to inspect acceptable products to ensure that identified foreign materials or portions thereof have not been missed by the rejection mechanism.
[0075] This invention can be used to report reliable foreign material identification to factories, and also to signal confirmation of rejection, or to issue additional alerts if foreign materials are still detected in the incoming product stream. Its value lies in food safety and assurance, confirmation of sorting machine functionality, reporting of quality control and product quality, and audit trails for foreign material detection. Advantages include reduced product recalls and reduced maintenance due to damage to processing machinery.
[0076] This invention can be used in conjunction with known methods and identifiers for foreign matter analysis. Aspects of artificial intelligence (AI) and the cloud can be incorporated into this invention. Known foreign matter (FM) alerts can be coupled with images of FMs, which can be stored and retrieved from the cloud. This invention focuses on confirming and reporting rejected FMs, and issuing warnings if rejection cannot be confirmed.
[0077] Figure 1 An embodiment of an apparatus according to an embodiment of the present invention for identifying at least one image of a foreign material from multiple images of suspected foreign materials from a sorting system is illustrated. An optical sorting system (1) is deployed in a product processing line to detect and remove unwanted foreign materials from incoming raw material products (2). The output of the sorting system is the sorted product (3) and the rejected foreign materials (4). The sorting system is designed and configured to transmit images of the detected foreign materials (7) to be rejected. These images are transmitted to an FM analyzer (5), which includes a computer, memory, storage devices, and neural network acceleration processing devices such as GPUs or FPGAs. The computer also includes multiple communication interfaces connecting the computer to at least one communication device (8) for images and for inference models for neural networks. The figure indicates that this storage is cloud-based and connected via the Internet, but... Figure 2 As shown, other options are possible. There are other possible options, and the location or connection of the storage library is not essential to the present invention. The computer (5) also includes a communication interface and components (9) for data and control, intended for connection to automation and control systems in the product processing line. The computer (5) includes a further communication interface and components (10) for signaling the true detection of foreign materials. This interface can convey simple status messages and, if necessary, images of the detected foreign materials.
[0078] Aside from the fact that the storage for FM images and the models for neural network inference engines are stored on local or remote computing and storage systems, rather than via an internet connection, this is also beneficial in situations where increased security levels are required or where the processing lines are independent. Figure 2 The diagram illustrates the relationship with Figure 1 The system shown is similar to that of the present invention. This configuration of the invention is limited and, for example, it is not as easy to generate and deploy new inference models based on image data from multiple sites.
[0079] Figure 3 An extension according to a further embodiment of the invention is shown. Figure 2The embodiment adds an inspection device (11) to the system, which generates images of rejected products in a format equivalent to the images (7) generated by the optical sorting system (1). These images of rejected objects (13) are compared by an image matching unit (12) with images (14) of FM rejected by the sorting machine sent from an FM analyzer (5). The image matching unit includes a computer, a memory, a storage device, and a neural network acceleration processing device. The computer also includes a communication interface for receiving images and for communicating whether a match has been found. Figure 3 The table in the output indicates the equipment's output, and the significance of any omissions identified may be serious enough to trigger a warning or even halt the production line. Figure 3 Not shown in the diagram, the output signal (15) may be connected back to the neural network FM analyzer (5) so that the results of image matching (12) can be uploaded to the processing line via the communication interface and medium (9). Note that the rejected FM image (14) may contain all such images (7) sent, or may contain only the true FM image determined by the FM analyzer.
[0080] According to a further embodiment of the present invention, Figure 4 Directly built on Figure 3 The embodiment illustrated includes an additional imaging device (16) on the receiving output end (3) of the optical sorter (1), which generates images of all objects accepted by the sorter. These images are transmitted to another image matching unit (18), the content of which is... Figure 3 Similar to the image matching unit (12) described herein, the functions of the two image matching units (13) and (18) can potentially be executed within an image processing system, processing images from each source serially or in parallel. The output (19) of this receiving-side image matching unit (18) can support the corresponding output (15) of the rejection-side unit. This signal can confirm correct rejection, confirm missed rejection, or raise concerns about product containment and the detectability of foreign materials if no foreign material is detected by either secondary inspection unit. The communications (15) and (19) can be connected back to the neural network FM analyzer (5) so that the results generated by the image matching (12) can be transmitted to the processing line via the communication interface and the medium (9).
[0081] Figure 3 and Figure 4The indication that the storage library (6) for images and models is local or remote is for illustrative purposes only, and the storage library (6) may also be cloud-based and connected via the Internet. Referring to the accompanying drawings, an apparatus for identifying images of foreign materials from a series of images of suspected foreign materials from a sorting system is further provided, comprising a sorting system (1) for detecting foreign materials from a stream of good products containing a number of foreign materials, a component (2) for transmitting images of the detected foreign materials, a computer (3) having a memory (4) and processing capabilities (5) suitable for image analysis (6) and a communication component (7) suitable for receiving the transmitted images (2), the computer (3) further comprising communication for signaling (8) the true identification (9) or false identification (10) of the foreign material based on the image analysis (6).
[0082] The image analysis process (6) may include an inference engine (11) and a machine learning model (12) trained on images of real and non-alien materials.
[0083] The image analysis process (6) may include image processing functions for recognizing shape (13), image processing functions for recognizing size (14), image processing functions for recognizing color (15), image processing functions for recognizing texture (16), and image processing functions for classifying materials using multivariate image analysis.
[0084] The image of the transmitted material (2) can be a multispectral or hyperspectral image with spectral bands in the range of 350 nm to 1700 nm.
[0085] The image obtained by true recognition (9) can be transmitted to a database and stored.
[0086] The communication (8) of the true identification (9) may include a signal that actuates the secondary rejection device.
[0087] The communication (8) and the image (2) representing true identification (9) can be communicated to the operator.
[0088] The computer (3) and processing power (5) may include at least one GPU (17) graphics processing unit or APU (18) acceleration processing unit for implementing the neural network associated with the machine learning model (12).
[0089] GPUs or APUs can be integrated into a central processing unit (CPU).
[0090] An apparatus for confirming the output of a sorting system (11) is also provided, comprising a component (12) for acquiring an image of a rejected object from the sorting system, a communication component (13) for acquiring an image of an accepted object from the sorting system, at least one camera and lighting device (14) for recording an image of a product rejected by the sorting system, optionally including at least one camera and lighting device (15) for recording an image of a product accepted by the sorting system, an image processing computer (16) for matching an image of an object intended to be rejected with an image of an object recorded by the camera (14) and confirming that it is not found in the image recorded by the camera (15), and a communication method for signaling the success or failure of the sorting rejection.
[0091] When used herein with reference to the present invention, the terms “comprises” and “having / including” are used to specify the presence of the stated features, integers, steps, or components, but do not exclude the presence or addition of one or more other features, integers, steps, components, or groups thereof.
[0092] It should be recognized that certain features of the invention described in the context of individual embodiments for clarity may also be provided in combination in a single embodiment. Conversely, various features of the invention described in the context of individual embodiments for brevity may also be provided individually or in any suitable sub-combination.
Claims
1. A method for identifying at least one image of a foreign material from a plurality of images of suspected foreign materials from a sorting system, the method comprising: Obtain at least one image of a suspected foreign material detected in a stream of good products containing a certain amount of foreign material; Perform image analysis on at least one image of the suspected foreign material; and Based on image analysis, suspected foreign materials are classified as either genuine foreign materials or non-foreign materials.
2. The method of claim 1, further comprising: When a suspected foreign material is classified as a genuine foreign material, a signal indicating that the suspected foreign material has been classified as a genuine foreign material is generated.
3. The method of claim 1, further comprising: When a suspected foreign material is classified as a non-foreign material, a signal indicating that the suspected foreign material has been classified as a non-foreign material is generated.
4. The method of any of the preceding claims, wherein performing image analysis includes using a machine learning model trained on images of both genuine and non-genuine materials.
5. The method of any of the preceding claims, wherein performing image analysis includes performing at least one of the following: a shape recognition processing function, a size recognition image processing function, a color recognition image processing function, a texture recognition image processing function, and a material classification function using multivariate image analysis.
6. The method of any of the preceding claims, wherein the at least one image of the dissimilar material is a multispectral or hyperspectral image with spectral bands in the range of 350 nm to 1700 nm.
7. The method of any of the preceding claims further includes storing an image of any suspected foreign material that is classified as a genuine foreign material.
8. The method of any one of claims 2-7, further comprising actuating a secondary rejection device using a signal indicating that a suspected foreign material has been classified as a genuine foreign material.
9. The method of any of the preceding claims further comprises providing at least one operator with an image of any suspected foreign material that is classified as a genuine foreign material.
10. An apparatus for identifying at least one image of a foreign material from a plurality of images of suspected foreign materials from a sorting system, the apparatus comprising: A component for obtaining at least one image of a suspected foreign material detected in a stream of good products containing a certain number of foreign materials; A component for performing image analysis on at least one image of suspected foreign material; and Components used for classifying suspected foreign materials as genuine foreign materials or non-foreign materials based on image analysis.
11. The apparatus of claim 10, further comprising a component for generating a signal indicating that the suspected foreign material has been classified as a genuine foreign material when the suspected foreign material is classified as a genuine foreign material.
12. The apparatus of claim 10, further comprising a component for generating a signal indicating that the suspected foreign material has been classified as a non-foreign material when the suspected foreign material is classified as a non-foreign material.
13. The apparatus of claim 10 or claim 11, wherein the component for performing image analysis includes a component for using a machine learning model trained on images of real and non-alien materials.
14. The apparatus of any one of claims 10 to 13, wherein the component for performing image analysis includes a component for performing at least one of: a shape recognition processing function, a size recognition image processing function, a color recognition image processing function, a texture recognition image processing function, and an image processing function for classifying materials using multivariate image analysis.
15. The apparatus of any one of claims 10 to 14, wherein the at least one image of the dissimilar material is a multispectral or hyperspectral image with spectral bands in the range of 350 nm to 1700 nm.
16. The apparatus of any one of claims 10 to 15, further comprising a component for storing an image of any suspected foreign material classified as a genuine foreign material.
17. The apparatus of any one of claims 11 to 16, further comprising a component for actuating a secondary rejection device using a signal indicating that a suspected foreign material has been classified as a genuine foreign material.
18. The apparatus of any one of claims 10 to 17, further comprising a component for providing at least one operator with an image of any suspected foreign material classified as a genuine foreign material.
19. A method for confirming the removal of foreign matter from a product stream during a sorting process, the method comprising: Obtain at least one image of at least one product removed from the product stream by the sorting system; Compare at least one image of at least one product to be removed with at least one image of foreign matter in the product flow that has been identified by the sorting system for removal; as well as When the image of at least one foreign object matches at least one image of at least one product removed from the product stream by the sorting system, it is determined that the foreign object has been successfully removed from the product stream.
20. The method of claim 19, wherein obtaining at least one image of at least one product removed from the product stream by the sorting system comprises recording said at least one image of the at least one product using a camera positioned along the rejection path leaving the sorting system.
21. The method of any one of claims 19 to 20, wherein obtaining at least one image of at least one product removed from the product stream by the sorting system comprises obtaining multiple images of multiple objects removed from the product stream.
22. The method of any one of claims 19 to 21, further comprising signaling confirmation of successful removal.
23. The method of any one of claims 19 to 22, further comprising obtaining at least one image of at least one product retained in the product stream by the sorting system.
24. The method of claim 23, further comprising: Compare at least one image of at least one retained product with at least one image of foreign matter in the product flow that has been identified by the sorting system for removal; as well as When the image of at least one foreign object matches at least one image of at least one product retained in the product stream by the sorting system, it is determined that the foreign object has not been successfully removed from the product stream.
25. The method of claim 24, further comprising signaling a determination that removal was unsuccessful.
26. The method of any one of claims 23 to 25, wherein obtaining at least one image of at least one product retained in the product stream by the sorting system comprises recording said at least one image of the at least one product using a camera positioned along the acceptance path leaving the sorting system.
27. A method for confirming the removal of foreign matter from a product stream during a sorting process, the method comprising: Obtain at least one image of at least one product removed from the product stream by the sorting system; The image of at least one product to be removed is compared with at least one image of foreign matter in the product flow that is identified by the sorting system for removal. as well as When the image of at least one foreign object does not match the image of at least one product removed from the product stream by the sorting system, it is determined that the foreign object was not successfully removed from the product stream.
28. The method of claim 27, wherein determining that the foreign object was not successfully removed from the product stream includes determining that the foreign object was not successfully removed from the product stream at the end of a time period.
29. The method of claim 27 or claim 28, wherein the time period begins at the time when the sorting system identifies foreign objects in the product stream.
30. An apparatus for confirming the removal of foreign matter from a product stream during a sorting process, the apparatus comprising: A component for obtaining at least one image of at least one product removed from the product stream by a sorting system; A component for comparing the at least one image of at least one product to be removed with at least one image of foreign matter in the product flow that is identified by the sorting system for removal; as well as The component used to determine that the foreign object has been successfully removed from the product stream when the at least one image of the foreign object matches at least one image of at least one product removed from the product stream by the sorting system.
31. The apparatus of claim 30, wherein the component for obtaining at least one image of at least one product removed from the product stream by the sorting system includes a component for recording said at least one image of the at least one product using a camera positioned along the rejection path leaving the sorting system.
32. The apparatus of any one of claims 30 to 31, wherein the component for obtaining at least one image of at least one product removed from the product stream by the sorting system includes a component for obtaining multiple images of multiple objects removed from the product stream.
33. The apparatus of any one of claims 30 to 32, further comprising a component for signaling confirmation of successful removal.
34. The apparatus of any one of claims 30 to 33, further comprising a component for obtaining at least one image of at least one product retained in the product stream by the sorting system.
35. The apparatus of claim 34, further comprising: A component for comparing the at least one image of at least one retained product with at least one image of foreign matter in the product flow that is identified by the sorting system for removal; as well as Used to determine a component from which the foreign object was not successfully removed from the product stream when the at least one image of the foreign object matches at least one image of at least one product retained in the product stream by the sorting system.
36. The apparatus of claim 35, further comprising a component for signaling a determination that was not successfully removed.
37. The apparatus of any one of claims 34 to 36, wherein the component for obtaining at least one image of at least one product retained in the product stream by the sorting system includes a component for recording said at least one image of the at least one product using a camera positioned along the acceptance path leaving the sorting system.
38. An apparatus for confirming the removal of foreign matter from a product stream during a sorting process, the apparatus comprising: A component for obtaining at least one image of at least one product removed from the product stream by a sorting system; A component for comparing the at least one image of at least one product to be removed with at least one image of foreign matter in the product flow that is identified by the sorting system for removal; as well as Used to determine a component from which a foreign object was not successfully removed from the product stream when the at least one image of the foreign object does not match the at least one image of at least one product removed from the product stream by the sorting system.
39. The apparatus of claim 38, wherein the component for determining that the foreign object was not successfully removed from the product stream includes a component for determining at the end of a time period that the foreign object was not successfully removed from the product stream.
40. The apparatus of claim 38 or claim 39, wherein the time period begins at the time when the sorting system identifies foreign objects in the product stream.
41. A computer-readable medium having instructions thereon, which, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-9 or 19-29.
42. A method generally as described herein with reference to any of Figures 1 to 5 or shown in Figures 1 to 5.
43. An apparatus generally as described herein with reference to any one of Figures 1 to 5 or shown in Figures 1 to 5.