Portable automatic device for determining impurities in andean grains using artificial vision

Abstract

The present invention relates to a portable automatic device for detecting impurities in Andean grains using artificial vision, which comprises a conveyor belt mounted on support rollers, a feed hopper, and an outlet tray for loading and collecting a sample. The device also includes a moveable tray coupled to an oscillating mechanism that allows the grains to be spread; a detection camera with a light source for capturing images under uniform lighting; a control and processing unit connected to a rechargeable battery, for analysing the images and quantifying the impurities using artificial vision algorithms; and a case for protecting and transporting the device.

Description

[0001] PORTABLE AUTOMATIC EQUIPMENT FOR THE DETERMINATION OF IMPURITIES IN ANDEAN GRAINS BY MACHINE VISION TECHNICAL FIELD

[0002] The present invention is applicable in the agro-industrial field for the quality control of Andean grains, more specifically it refers to a device for determining impurities in Andean grains by means of artificial vision.

[0003] BACKGROUND OF THE INVENTION

[0004] Currently, the prevailing methodology for analyzing impurities in Andean grains is largely based on a manual approach, in which a human operator performs the inspection. The protocol consists of sample preparation, visual evaluation, identification and categorization of impurities, followed by recording and documenting the findings. However, this approach has limitations that affect its effectiveness, such as accuracy, consistency, and efficiency.

[0005] Patent application CN210090320U describes an image acquisition device for detecting weed seeds comprising a translucent conveyor belt, a camera positioned above the conveyor belt, and a light source located below the conveyor belt. However, the light source below the conveyor belt can cast shadows or create uneven illumination on the seeds, complicating image processing and accurate detection. Furthermore, the transparency of the belt could cause light refraction or distortion, affecting image capture accuracy, especially if the seeds are moving rapidly.

[0006] Patent CN214794500U discloses a detection device that enables real-time, online measurement of the external appearance quality of wheat. The device comprises a conveyor belt, a hopper, a single grain guide slot, a dark box, a linear array light source, an image acquisition device, a display screen, and a computer. However, the guide slot requires the grains to be perfectly aligned for accurate detection. If the grains are not properly aligned, the image acquisition device may capture incomplete or distorted images, affecting the measurement accuracy.

[0007] On the other hand, application ES1057192U provides a machine vision system for inspecting packaged food products. This system consists of a conveyor line for unsealed packages and a support structure to which a camera connected to a computer with an image capture card is attached. However, the inclusion of the computer, which would require a monitor, keyboard, and other accessories, makes the system occupy considerable space in the production environment.

[0008] Finally, the website https: / / repositoho.upeu.edu.pe / handle / 20.500.12840 / 1945 describes a device and application system that captures images to analyze the macroscopic impurities of quinoa. The analysis procedure consists of placing the samples in quadrants, considering that each quadrant represents an image and each image must contain 9 units comprising both quinoa grains and possible impurities. This implies a prolonged use of time in the ordering of the samples by an operator.

[0009] DESCRIPTION OF THE INVENTION

[0010] In order to solve the aforementioned problems, an automatic device was developed that allows the detection of impurities in Andean grains as part of the quality control process using artificial vision.

[0011] The equipment comprises a conveyor belt attached to support rollers with a feed hopper; a movable tray attached to support shafts arranged below the hopper to spread the sample onto the conveyor belt; a light source with a high-speed camera that processes images to detect and quantify impurities in Andean grains; and an output tray with a storage unit.

[0012] The inclusion of the movable tray allows for a back-and-forth motion, providing a dispersion system. This gives the present invention the advantage of spreading the sample on the conveyor belt and ensuring that the grains are arranged in a single-layer bed, thus preventing overlap and facilitating the task of machine vision algorithms. Therefore, it even improves the evaluation of samples with a high proportion of small or irregularly shaped grains.

[0013] For the reasons mentioned above, the invention is distinguished from its predecessors by the simplicity of its use, since it only requires placing the sample directly into the feed hopper, and its arrangement is not necessary due to the inclusion of the movable tray.

[0014] It is worth mentioning that the light source is located at the top of the conveyor belt, ensuring uniform lighting and preventing the formation of shadows, which is essential to obtain high-quality images.

[0015] The images are captured in high-resolution video which are subsequently processed by machine vision algorithms, allowing the accurate detection of a wide range of impurities, from small fragments of foreign matter to damaged or smaller grains.

[0016] Additionally, it comprises a control and processing unit containing a processor and a display screen equipped with batteries that provide extended autonomy, so the invention is characterized as a portable and easy-to-move device, allowing its use for a significant time without the need for an electrical connection.

[0017] For the reasons mentioned above, this feature makes it an ideal tool for grain inspection in the field, in collection centers or in small and medium-sized processing companies.

[0018] BRIEF DESCRIPTION OF THE FIGURES

[0019] The description of the present invention will be further expanded upon with reference to the selected embodiments shown in the accompanying figures, in which:

[0020] Figure 1: Isometric view of the portable automatic equipment for the determination of impurities in Andean grains by artificial vision, showing the protective cubicle for the main components and the arrangement of a handle for transporting the equipment.

[0021] Figure 2: Isometric view of the internal parts of the portable automatic equipment of Figure 1, showing the arrangement of a hopper, a conveyor belt and an output tray with a system to detect impurities in the grain sample by means of a detection camera (7).

[0022] Figure 3: Opposite perspective view of Figure 2, showing the elements of the conveyor belt and the moving tray.

[0023] Figure 4: Detailed view of Figure 3, showing the pivot axis for the movable tray.

[0024] PREFERRED EMBODIMENTS OF THE INVENTION

[0025] The attached figures illustrate an example of an embodiment of portable automatic equipment for detecting impurities in Andean grains by means of machine vision; it should be understood that the drawings and description are not intended to limit the broad aspects of the invention to the embodiments illustrated.

[0026] Said equipment comprises a conveyor belt (1) with a feed hopper (3) for loading the sample to be analyzed; a movable tray (4) arranged under the feed hopper (3) to form a single-layer bed; and an output tray (9) for collecting the analyzed sample.

[0027] Additionally, it comprises an oscillating movement mechanism (12) coupled to the movable tray (4) for spreading the grain load onto the conveyor belt (1); a detection camera (7) with a light source (6) mounted on top of the conveyor belt (1) for capturing images of the grain sample; a control and processing unit (8) configured to process the images captured by a detection camera (7); and a rechargeable battery (10) connected to the control and processing unit (8), which provides operating autonomy to the portable automatic equipment for detecting impurities.

[0028] Figure 2 shows that the conveyor belt (1) is supported at each end by support rollers (2), which are designed to transport the grains through the inspection area; also at one end of the conveyor belt (1), a support structure (5) is provided for mounting the feed hopper (3) and securing the movable tray (4) which is attached to the oscillating movement mechanism (12).

[0029] Figure 3 shows that the light source (6) is arranged at the level of the detection camera (7), both mounted on a tubular configuration stand (6a) with two adjustable mounting brackets, which are adjusted with respect to the floor level to a sufficient height; allowing the light source (6) and the detection camera (7) to be placed at a higher level with respect to the conveyor belt (1), achieving uniform illumination for the capture of images of the grain sample to be analyzed.

[0030] According to the example shown, the light source (6) consists of an LED ring, in the center of which is the high-speed detection camera (7). This configuration ensures uniform illumination, eliminating shadow formation during image capture of the grain sample.

[0031] Furthermore, the control and processing unit (8) comprises a processor (8a) and a display screen (8b), as illustrated in Figure 1. These are configured to process and analyze the images captured in real time by the detection camera (7) using machine vision algorithms, in order to then detect and quantify impurities in the grain sample of different sizes and types. The aforementioned control and processing unit (8) is configured to control the folding of the movable tray (4) and form a single-layer bed of grains, allowing the images captured by the detection camera (7) to be processed to detect and quantify impurities in the Andean grain sample using machine vision algorithms based on convolutional neural networks (CNNs).Additionally, this figure illustrates the oscillating motion mechanism (12) responsible for generating the back-and-forth motion of the mobile tray (4) and spreading the grain load on the conveyor belt (1) to form a bed; which consists of a motor (12a), a flywheel with an eccentric shaft (12b) and a connecting rod (12c), which provides the oscillating motion to the mobile tray (4).

[0032] Figure 4 shows that the support structure (5) includes, at its rear end, a square support (5a) for mounting the feed hopper (3), and at the opposite end, a shaft (5b) for coupling and folding the movable tray (4) by means of a linear guide (4a), and also features a rectangular body (4b) that is fixed to the oscillating movement mechanism (12).

[0033] Additionally, Figure 1 illustrates that the portable automatic equipment comprises a cubicle (11) of quadrangular configuration to contain all the pieces of the equipment that fulfills the function of protection and containment.

[0034] The cubicle (11) has, at the top, a feed gate (11 a) leading to the feed hopper (3) for protection, and a handle (11 b) for transporting the portable equipment; while at the lower front, it has an output gate (11c) for removing the output tray (9); and at the base, safety feet (11 d) and also a power button (11 e).

[0035] For the handling of electronic devices, the cubicle (11) also has, on its surface, slots for mounting the configuration panel of the detection camera (7) and the display screen (8b) of the control and processing unit (8) for monitoring.

Claims

CLAIMS 1. A portable automatic device for determining impurities in Andean grains using machine vision, consisting of a conveyor belt (1) fixed with support rollers (2), and a feed hopper (3) with an output tray (9) for loading grains and collecting the analyzed sample, characterized in that it comprises: - a movable tray (4) arranged below the feed hopper (3) to form a single-layer bed on the conveyor belt (1); - an oscillating motion mechanism (12) to generate the back-and-forth motion of the movable tray (4) and spread the grain load onto the conveyor belt (1); - a detection camera (7) mounted on top of the conveyor belt (1) to capture images of the grain sample; - a light source (6) arranged at the level of the detection camera (7) to provide uniform illumination for image capture; - a control and processing unit (8) configured to process the images captured by the detection camera (7), as well as to detect and quantify impurities in the grain sample using machine vision algorithms; - a rechargeable battery (10) connected to the control and processing unit (8); and - a cubicle (11) featuring a feed door (11 a) and a handle (11 b) for the protection and transport of the equipment.

2. A portable automatic equipment for the determination of impurities in Andean grains by means of artificial vision, according to claim 1, characterized in that it comprises a support structure (5) arranged on the conveyor belt (1) for mounting the feed hopper (3) and securing the movable tray (4) with the oscillating movement mechanism (12).

3. A portable automatic equipment for the determination of impurities in Andean grains by means of artificial vision, according to claim 2, characterized in that the support structure (5) includes, at its rear end, a quadrangular support (5a) for mounting the feed hopper (3), and at the opposite end, a shaft (5b) for coupling and folding the movable tray (4).

4. A portable automatic equipment for the determination of impurities in Andean grains by means of artificial vision, according to the previous claims, characterized in that the movable tray (4) has a linear guide (4a) to couple to the axis (5b) of the support structure (5) and a rectangular body (4b) that is fixed to the oscillating movement mechanism (12), which includes a motor (12a), a flywheel with eccentric axis (12b) and a connecting rod (12c).

5. A portable automatic equipment for the determination of impurities in Andean grains by means of artificial vision, according to claim 1, characterized in that the detection camera (7) with the control and processing unit (8) integrate an artificial vision system configured to process the captured images in real time.

6. A portable automatic device for determining impurities in Andean grains by means of machine vision, according to the preceding claims, characterized in that it comprises a stand (6a) for mounting the light source (6) and the detection camera (7) at a sufficient height to capture images of the grain sample with uniform illumination.

7. A portable automatic equipment for the determination of impurities in Andean grains by means of artificial vision, according to claims 1 and 3, characterized in that the control and processing unit (8) is constituted by a processor (8a) and a display screen (8b) configured to analyze in real time the images captured by the detection camera (7) and detect impurities in the grain sample of different sizes and types.

8. A portable automatic device for determining impurities in Andean grains by means of machine vision, according to claim 1, characterized in that the cubicle (11) has, on the lower front, an outlet gate (11c) for extracting the outlet tray (9), and on the upper part, a feed gate (11a) to cover the feed hopper (3).

9. A portable automatic device for determining impurities in Andean grains by means of machine vision, according to claims 1 and 8, characterized in that the cubicle (11) has, on its surface, the configuration panel of the detection camera (7) and the display screen (8b) of the control and processing unit (8) for monitoring.

10. A portable automatic device for determining impurities in Andean grains by means of artificial vision, according to the previous claims, characterized in that the cubicle (11) has, at its base, safety feet (11 d) and also a power button (11 e).

Images