Patent Application
20230213460
The present invention relates to a system for determining the number of small products in the bag package from the x-ray image that enables detecting the missing quantities of small products or packages in a disorganized and random position in the bag package.
In particular, the present invention relates to the ability to sort the bag packages with missing numbers at the x-ray output by estimating the number of small packages in the bulk at random positions in three dimensions in the bag package.
Today, conveyor systems that perform weighing in the product transition are used to provide the number of products inside the bag pack. The weighing process is performed as the product passes through the product transport conveyor working on the load cell called Loadcell. Depending on the fluctuation in weight of the product, missing pieces can be detected by this method.
In food production lines, there are methods developed to verify the in-pack quantity from the images obtained as the bag-packed products pass through the x-ray scanner. However, in the current x-ray image processing algorithms, the missing pieces of small products or packages that are disorganized and in random positions in the bag package cannot be detected.
The invention, which is the subject of the application numbered “JP2002214357A” in the state of the art, relates to an X-ray image inspection system and method capable of managing the quality of products according to numbers by which quality information about contamination with foreign matter and the number of contents is imparted to the products. A metal detector is used to check whether or not the products contain iron as a foreign matter. A weighing instrument measures the weight per one product. X-ray images of the products are taken and stored in an X-ray foreign matter inspection apparatus. A laser marker marks the package of the product by using a laser with a mark including data about the date of inspection or fabrication of the product.
The invention, which is the subject of the application numbered “JP2012242289A” in the state of the art, is a method for inspecting the number of products and package contents with high precision inspection in order to reduce the need for workers. Packaged products on the vertically moving band are inspected one by one by X-ray imaging, provided that they do not superimpose. In the image obtained after X-ray imaging, the number of products is determined by counting the protrusions and corners of the packaged product.
In the state of the art, there are systems that perform imaging for the package with X-ray and enable sorting the package in which the defective product is detected. In X-ray image processing algorithms, there is a need for a system that detects the missing pieces of small products or packages that are disorganized and in random positions in the bag package.
Consequently, the disadvantages disclosed above and the inadequacy of available solutions in this regard necessitated making an improvement in the relevant technical field.
The present invention relates to a system for determining the number of small products in the bag package from the x-ray image that enables detecting the missing quantities of small products or packages in a disorganized and random position in the bag package.
The most important object of the present invention is to detect the missing quantities of small products or packages in disorganized and random positions.
Another object of the present invention is to enable the number of products to be determined by the contrast difference of the images.
Another object of the present invention is to ensure that the products that are found to be missing products are sorted out.
The present invention relates to a system for determining the number of small products in the bag package from the x-ray image that enables detecting the missing quantities of small products or packages in a disorganized and random position in the bag package.
The system for determining quantities of small products in the bag package from the x-ray image (100) comprises bag package (101), vibrating band (102), vibration unit (103), input conveyor (104), main conveyor (105), x-ray carrier (106), x-ray detector (107), output conveyor (108), control unit (109), image processing unit (110), compressed air valve (111), and input-output card (112).
The bag package (101) contains the disorganized, free, and tangled products.
The vibrating band (102) prevents the products in the bag package (101) from clumping together with the vibration as it passes over the bag package (101), thereby making it more discrete and visually distinguishable.
The vibration unit (103) generates vibration on the vibrating band (102) and enables its amplitude and intensity to be adjusted.
The input conveyor (104) provides the transport of the bag package (101) coming from the vibrating band (102) to the main conveyor (105).
The main conveyor (105) carries the x-ray scanner (106) and x-ray detector (107) and the control unit (109) thereon, allowing the bag package (101) to be conveyed to the output conveyor (108).
The X-ray scanner (106) provides scanning of the products in the bag package (101) moving on the main conveyor (105).
The X-ray detector (107) creates the numerical data of the rays passing through the x-ray scanner (106) according to the energy level and provides creating a two-dimensional image.
The control unit (109) receives the two-dimensional images from the x-ray sensor (107) and ensures that they are saved to the computer therein. The control unit (109) provides the transfer of the images recorded to the image processing unit (110) via the ethernet connection (113).
The image processing unit (110) enables determining the parts of the product of interest in the package in the two-dimensional images receiving from the processor and the control unit (109), by filtering. The image processing unit (110) provides the filtering of parts with pixel value between 0 and 200 (adjusted) numerically by filtering. Thus, the product regions are detected. Upper and lower limit values of filter can be adjusted by the user. The image processing unit (110) determines the product groups and their limits that are not adjacent to each other from the filtered sections. The image processing unit (110) numerically calculates the grayness values for each group. The image processing unit (110) calculates the total grayness value by summing the grayness value in each pixel. The image processing unit (110) ensures that the moving average value of the total grayness values of each package is obtained, and this average value is divided by the number of in-pack units being worked at that time in order to obtain the average grayness value per product. The moving average number is a user changeable value. By means of the moving average, the mass change and indirectly the grayness change caused by the food processing differences of the product in the currently running process are changed in an adaptive manner, thereby resulting in more accurate results. The image processing unit (110) provides the estimation of how many products each group belongs to by dividing the calculated grayness value of each group by the average grayness per product unit value. Thus, the total number of items in the package is found. In another embodiment of the present invention, the image processing unit (110) provides the estimation of how many products there are by dividing the total grayness numerical value of the package by the unit product grayness value. The resulting product estimate value is a real number. The image processing unit (110) limits the sorting limit values of the package over real values.
The output conveyor (108) enables transporting the bag package (101) coming out of the main conveyor (105).
The input output card (112) provides the commands from the image processing unit (110) to be transmitted to the compressed air valve (111).
The compressed air valve (111) ensures that the defective ones from the bag packages (101) moving on the output conveyor (108) are sorted with the help of compressed air.
In an experimental study performed on an x-ray machine, 3-layer wafer slices covered with cream were scanned individually from x-ray. The measured gram mass values of each scanned product are different. This difference is due to differences in cream amount and other non-ideal processes. The pixel area value and the total grayness value were calculated by processing the image of each product and adapted to the table.
In Table 1, the gray area area value (AREA), total grayness value (GRAY) and weight (GRAM) values for each product obtained from the grayscale image are given. And also (GRAY/GRAM) and (AREA/GRAM) values are also derived for analysis. With the examination, the lowest deviation value from these values is the GRAY/GRAM value. This shows that the data that is compatible with the weight (GRAM) value of the products is the total value of grayness (GRAY).
The data obtained from performing the same experiment with different products were shown in Table 2. Here, as a result, it is observed that the sum of the gray values in the sections of the product in the images obtained from x-ray gives proportional data with the total mass.
A method for determining (100) the number of small product in a bag package from the X-ray image, comprising;
| Number | Date | Country | Kind |
|---|---|---|---|
| 2022/000080 | Jan 2022 | TR | national |
