WO2024109669A1 - 一种检测包装盒内物料放置情况的方法 - Google Patents

一种检测包装盒内物料放置情况的方法 Download PDF

Info

Publication number
WO2024109669A1
WO2024109669A1 PCT/CN2023/132500 CN2023132500W WO2024109669A1 WO 2024109669 A1 WO2024109669 A1 WO 2024109669A1 CN 2023132500 W CN2023132500 W CN 2023132500W WO 2024109669 A1 WO2024109669 A1 WO 2024109669A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
value
grayscale
area
original image
Prior art date
Application number
PCT/CN2023/132500
Other languages
English (en)
French (fr)
Inventor
瞿诗雄
姜健军
Original Assignee
利多(香港)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 利多(香港)有限公司 filed Critical 利多(香港)有限公司
Publication of WO2024109669A1 publication Critical patent/WO2024109669A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Definitions

  • the present invention relates to the technical field of product packaging, and more particularly to a method for detecting the placement of materials in a packaging box.
  • the existing devices on the market currently require two independent machine vision systems or an industrial computer with two cameras.
  • Two independent machine vision systems or an industrial computer with two cameras are used to detect the presence of cotton swabs and certificates.
  • the first machine vision system is placed after the card issuing machine (certificate feeding box), and uses template matching to detect the presence of the certificate based on the specific pattern features on the front of the certificate (for example, detecting whether there is a circle on the certificate in the picture).
  • the second machine vision system is placed after the cotton swab spider hand (cotton swab feeding box), and detects whether there is strong reflection in the box (the back of the cotton swab is facing up at this time, and the cotton swab can be seen), or uses template matching (the front of the cotton swab is facing up at this time, and the specific image features on the front of the cotton swab, such as the rectangular black mark in the corner) are used to detect the presence of the cotton swab.
  • the present invention overcomes the deficiencies of the prior art and provides a packaging method with simple design, low cost and high product adaptability.
  • the technical solution of the present invention is as follows.
  • the present invention provides a method for detecting the placement of materials in a packaging box, comprising the following steps:
  • Original image acquisition step obtaining a grayscale image of the material placement area in the packaging box, selecting a target area to be analyzed, and obtaining a grayscale image of the target area as the original image A;
  • Image reduction step The original image A in step 1) is reduced in proportion to obtain image B;
  • Image enlargement step enlarge the image B in step 2) to the size of the original image A, and obtain an image C of the same size as the original image A;
  • Image processing step subtract image C from original image A to obtain image D, multiply image D by a corresponding coefficient or multiply image D by a corresponding coefficient plus a certain value to obtain image E;
  • Confirmation step Calculate the variance of the grayscale value of the target area according to image E to obtain the variance value Z. If the variance value Z is within the set interval, it is determined that the material in the packaging box meets the requirements; if the variance value Z is not within the set interval, it is determined that the material in the packaging box does not meet the requirements.
  • step 1) also includes the step of binarizing the original image A, comparing the binarized image with a threshold value, and if the grayscale value of the binarized image is less than the threshold value, it is determined that the original image A has no highlight area, and the process proceeds to step 2); if there is a part in the binarized image with a grayscale value greater than the threshold value, there is a highlight part, and the original image A needs to be binarized to obtain the highlight area F, which is then expanded to obtain the area F+, and a value is assigned to F+.
  • step 1) a fixed grayscale value is assigned to the pixels in the F+ area, and the image is used as the original image A to enter step 2), or in step 4), a fixed grayscale value is assigned to the pixels in the F+ area in the image E, and then entering step 5).
  • the assigned value is 0 or a low brightness value.
  • step 4 D is multiplied by a corresponding coefficient or the image D is multiplied by a corresponding coefficient plus a certain value. If the grayscale value of a pixel of the image D exceeds 255 after the above calculation, the pixel is set to 255.
  • the corresponding coefficient in step 4) is a value greater than 0, and the certain value is a positive number or a negative number.
  • the present invention also provides a packaging method, including providing a packaging device, the packaging device including a switching power supply, a visual controller, a rejection device, a conveyor belt for conveying packaging boxes, and a camera located above the conveyor belt, the visual controller is electrically connected to the camera to receive the internal image of the packaging box taken by the camera, the visual controller determines whether the material in the packaging box meets the regulations, if it is determined that the material in the packaging box does not meet the regulations, the visual controller outputs a signal, the rejection device is started, and the unqualified packaging box is rejected from the conveyor belt, the step of determining whether the material in the packaging box meets the regulations includes:
  • Original image acquisition step Use the camera on the packaging equipment to obtain a grayscale image of the material placement area in the packaging box on the conveyor belt, select the target area to be analyzed, and obtain the grayscale image of the target area as the original image A;
  • Image reduction step The original image A in step 1) is reduced in proportion to obtain image B;
  • Image enlargement step enlarge the image B in step 2) to the size of the original image A, and obtain an image C of the same size as the original image A;
  • Image processing step subtract image C from original image A to obtain image D, multiply image D by a corresponding coefficient or multiply image D by a corresponding coefficient plus a certain value to obtain image E;
  • Confirmation step Calculate the variance of the grayscale value of the target area according to image E to obtain the variance value Z. If the variance value Z is within the set interval, it is determined that the material in the packaging box meets the requirements; if the variance value Z is not within the set interval, it is determined that the material in the packaging box does not meet the requirements.
  • the packaging device also includes a trigger sensor for sensing that the packaging box enters the camera shooting area, and the trigger sensor is electrically connected to the camera.
  • the trigger sensor is a photoelectric switch.
  • the packaging equipment also includes a removal opening.
  • step 1) also includes a binarization step of the original image A, comparing the binarized image with the threshold value, and if the grayscale value of the binarized image is less than the threshold value, it is determined that the original image A has no highlight area, and the process proceeds to step 2); if there is a part of the binarized image with a grayscale value greater than the threshold value, there is a highlight part, and the original image A needs to be binarized to obtain the highlight area F, which is then expanded to obtain the area F+.
  • the pixels in the F+ area can be assigned a fixed grayscale value, and the image can be used as the original image A to proceed to step 2).
  • step 4 the pixels in the F+ area of image E can be assigned a fixed grayscale value, and then the process proceeds to step 5).
  • step 5 the step of assigning a fixed grayscale value to the pixels in the F+ area can be selected only one step or both.
  • the corresponding coefficient in step 4) is a value greater than 0, and the certain value is a positive number or a negative number.
  • the present invention has the following advantages: the present invention can replace manual work through optimized design. Moreover, the hardware equipment only has one machine vision system, which further reduces the cost and greatly reduces the floor space of the equipment, which is conducive to the development of more production lines.
  • the operator of the present invention can change the values in the file in real time according to the new material to adapt to the new product. Small local pattern changes do not affect the operation of the visual system, and the compatibility is strong.
  • Figure 1 is an image of a packaging box without any material stored in it.
  • Figure 2 shows an image of the certificate of conformity placed face up in the packaging box.
  • Figure 3 shows an image of a cotton swab and a certificate of conformity placed in a packaging box, with the paper side of the cotton swab facing upward.
  • Figure 4 shows an image of the certificate of conformity placed face down in the packaging box.
  • FIG. 5 is an image of a cotton swab package in a packaging box with the plastic side facing upward.
  • FIG. 6 is a schematic diagram of a flow chart of determining the material mode of a packaging box according to the present invention.
  • FIG. 7 is a schematic diagram of a device configuration circuit of the present invention.
  • FIG. 8 is a schematic diagram of the packaging device of the present invention.
  • FIG. 9 is a schematic diagram of the packaging device of the present invention from another angle.
  • the packaging device 100 shown in FIG. 7 , FIG. 8 and FIG. 9 includes a switch power supply of the device, a conveyor belt 101 for conveying packaging boxes, packaging boxes 200 are spaced a certain distance from each other and arranged in sequence on the conveyor belt 101, and a camera 102 and a camera light source 103 are respectively mounted on a bracket 106.
  • the camera 102 and the camera light source 103 are mounted just above the conveyor belt 101, and the camera light source 103 is located between the camera and the conveyor belt 101.
  • the camera 102 is directly mounted on the bracket 106, and the camera light source 103 is connected to the bracket 106 through a light source bracket 107.
  • the camera light source 103 uses a circular ring light, and a shooting through hole 108 is provided on the light source bracket 107.
  • the camera 102 is aimed at the inside of the packaging box through the shooting through hole 108 and the central unlit area of the circular ring light to shoot.
  • the trigger sensor 104 is mounted on the mounting bracket 109 of the conveyor belt 101.
  • the trigger sensor 104 (SQ1 in FIG. 7 ) senses that the package box has entered the camera shooting area when the package box is transported to the trigger sensor.
  • the trigger signal (the photoelectric switch can be set to rising edge trigger, falling edge trigger, connection trigger, etc.) directly triggers the camera to take pictures through the circuit.
  • the visual controller receives the internal pictures of the package box taken by the camera during the timed scanning.
  • the visual controller judges whether the feeding in the package box meets the regulations according to the method shown in the flowchart of judging the feeding situation of the package box shown in FIG. 6. If it is judged to meet the regulations, the qualified package box continues to move forward along the conveyor belt. If it is judged that the feeding does not meet the regulations, for example, the package box lacks materials or the material direction is reversed, the output point of the visual controller (shown as 1 of the visual controller in FIG. 7 ) has an output, thereby connecting the coil of the intermediate relay KA1, the normally open contact of KA1 is closed, and the rejection solenoid valve YV1 is turned on. The rejection device of the packaging equipment pushes the unqualified package box into the package box rejection opening 105. The reject device and the carton reject opening are located downstream of the trigger sensor.
  • the camera light source can also be installed at other locations of the packaging equipment; or the camera light source may not be installed while ensuring the brightness required for taking pictures; or the camera light source can be directly mounted on the camera.
  • the trigger sensor can also be installed at other locations of the packaging equipment, and other components that can determine that the packaging box has entered the shooting area can be used.
  • the trigger sensor can also be not installed, and the camera takes pictures continuously.
  • the rejection device is activated to export the unqualified packaging box.
  • the rejection device uses a rejection solenoid valve. When the rejection is started, the air pipe of the rejection device will be connected, and the unqualified packaging boxes will be blown out of the conveyor belt by air jet.
  • the rejection device can also be used in other ways.
  • the following is a packaging method of the present invention for judging whether the placement of materials in a packaging box complies with regulations, specifically comprising the following steps:
  • Original image acquisition step obtain a grayscale image of the material placement area in the packaging box, select a target area to be analyzed, and obtain a grayscale image of the target area as the original image A.
  • the target area can be the entire image area in the packaging box or a certain area within the entire image area.
  • Image reduction step The original image A is reduced in proportion to obtain a reduced image B.
  • Image enlargement step enlarge image B to the size of the original image A, and obtain image C of the same size as the original image A.
  • Image processing step subtract image C from original image A to obtain image D, multiply image D by a corresponding coefficient or multiply image D by a corresponding coefficient plus a certain value to obtain image E.
  • the corresponding coefficient is a value greater than 0, and the certain value can be a positive number or a negative number.
  • the corresponding coefficient and the certain value can be determined based on tests.
  • Confirmation step Calculate the variance of the grayscale value of the target area according to the image E to obtain the variance value Z. If the variance value Z is within the set interval, the placement of the materials in the packaging box meets the requirements and is confirmed as a qualified product; if the variance value Z is not within the set interval, the placement of the materials in the packaging box does not meet the requirements and is confirmed as a defective product.
  • the following is a packaging method of the present invention for judging whether the placement of materials in a packaging box complies with regulations.
  • This method is a processing method for when there is a highlight part in the image. Compared with method 1, it is different from method 1 in that after obtaining the original image A, it is necessary to judge whether there is a highlight area and process the highlight area. Specifically, it includes the following steps:
  • Original image acquisition step obtain a grayscale image of the material placement area in the packaging box, select a target area to be analyzed, and obtain a grayscale image of the target area as the original image A.
  • the target area can be the entire image area in the packaging box or a certain area within the entire image area.
  • step 2) if it is determined in step 2) that the original image A has no highlight area, the original image A is reduced in proportion to obtain a reduced image B; if it is determined in step 2) that the original image A has a highlight area, the image after dilation processing of the highlight area is reduced in proportion to obtain a reduced image B.
  • Image enlargement step enlarge image B to the size of the original image A, and obtain image C of the same size as the original image A.
  • Image processing step image D is obtained by subtracting image C from original image A, and image D is multiplied by a corresponding coefficient or image D is multiplied by a corresponding coefficient plus a certain value to obtain image E.
  • the corresponding coefficient is a value greater than 0, and the certain value can be a positive number or a negative number.
  • the corresponding coefficient and the certain value can be determined based on tests.
  • Confirmation step Calculate the variance of the grayscale value of the target area according to the image E to obtain the variance value Z. If the variance value Z is within the set interval, the placement of the materials in the packaging box meets the requirements and is confirmed as a qualified product; if the variance value Z is not within the set interval, the placement of the materials in the packaging box does not meet the requirements and is confirmed as a defective product.
  • Method 2 also includes the step of assigning a value to the region F+.
  • the assignment step can be performed in step 2), that is, the region F of the highlight part is expanded to obtain the region F+, and the grayscale value of the region F+ of the original image A is assigned to 0 or a low brightness value.
  • the assignment step can also be performed in step 5), that is, the grayscale value of the region F+ in the image E is assigned to 0 or a low brightness value.
  • FIG1 is an image of a packaging box without the two materials, cotton swabs and certificates of conformity, wherein image a in FIG1 is not processed by the method of the present invention, but is directly obtained by photographing the material placement area in the packaging box, and its grayscale value variance is 22.3106; image b in FIG1 is an image processed by method 1 of the present invention, and its grayscale value variance is 44.8119.
  • Figure 2 is an image of a certificate placed in a packaging box, wherein image a in Figure 2 is not processed by the method of the present invention, but is a grayscale image directly obtained by photographing the material placement area in the packaging box, and its grayscale value variance is 31.3447, and image b in Figure 2 is an image processed by method 1 of the present invention, and its grayscale value variance is 87.2330.
  • Figure 3 is an image of a cotton swab and a certificate of conformity placed in a packaging box, wherein image a in Figure 3 is not processed by the method of the present invention, but is a grayscale image directly obtained by photographing the material placement area in the packaging box, and its grayscale value variance is 35.8878.
  • Image b in Figure 3 is an image processed by method 2 of the present invention, and its grayscale value variance is 63.4496.
  • the gray value variance of the image with only the certificate of conformity is different from that of the image without the certificate of conformity.
  • the difference between the image grayscale value variance of the cotton swab and the certificate is 42.4211 (87.2330-44.8119), and the difference between the image grayscale value variance of the cotton swab and the certificate is 23.7834 (87.2330-63.4496).
  • the grayscale value variances of these three cases are very different, and the image measurement deviation will not have a significant impact on the image grayscale value variance. Therefore, the packaging equipment can accurately determine whether the correct material is placed in the packaging box.
  • the packaging method of the present invention can effectively and accurately determine whether the correct material is placed in the packaging box.
  • Example 2 Determining whether the placement direction of the materials in the packaging box is correct
  • the front side of the certificate is placed in the packaging box with the front side facing up
  • Figure 4 is placed in the packaging box with the front side facing down.
  • the front side of the certificate refers to the side with text
  • the back side of the certificate refers to the side without text.
  • the grayscale image in the packaging box is obtained, such as image a in Figure 2 and image a in Figure 4.
  • the image a is processed by method 1 of the present invention, and the variance of the grayscale image of image b in Figure 2 is 87.2330, and the variance of the grayscale image of image b in Figure 4 is 42.325.
  • the variance value Z of the grayscale image is greater than the set value 66.
  • the variance of the grayscale image of image b in Figure 2 is 87.2330, which is greater than the set value of 66.
  • the packaging machine determines that the product certificate in the packaging box is placed face up, which is correct placement.
  • the variance of the grayscale image of image b in Figure 4 is 42.325, which is less than the set value of 66.
  • the packaging machine determines that the product certificate in the packaging box is placed face down, which is incorrect placement and needs to be removed from the packaging line.
  • Example 3 Determining whether there is material in the packaging box
  • the grayscale image in the packaging box is obtained, such as the a image in Figure 1 and the a image in Figure 2.
  • the a image is processed by the method 1 of the present invention, and the variance of the grayscale image of the b image in Figure 1 is 44.8119, and the variance of the grayscale image of the b image in Figure 2 is 87.2330.
  • the variance value Z of the grayscale image is greater than the set value 60.
  • the variance of the grayscale image of the b image in Figure 1 is 44.8119, which is less than the set value 60.
  • the packaging machine determines that there is no material in the packaging box, it is an empty box, it is wrongly placed, and needs to be removed from the packaging line.
  • the variance of the grayscale image of the b image in Figure 2 is 87.2330, which is greater than the set value 60.
  • the packaging machine determines that the material has been placed in the packaging box, which is correctly placed.
  • Example 4 Determining whether the quantity and type of materials in the packaging box are correct
  • FIG 1 there is no material in the packaging box, that is, no product certificate is placed in it.
  • Figure 2 shows that a single material is placed in the packaging box, that is, the certificate is placed in it.
  • Figure 3 shows that two materials are placed in the packaging box, that is, the certificate and the cotton swab are placed in it. Placement completed After that, the grayscale images in the packaging box are obtained, such as the image a in FIG1, the image a in FIG2, and the image a in FIG3.
  • the image a is processed by the method 1 of the present invention, and the variance of the grayscale image of the image b in FIG1 is 44.8119, the variance of the grayscale image of the image b in FIG2 is 87.2330, and the variance of the grayscale image of the image b in FIG3 is 63.4496.
  • the variance value Z is set within 49 ⁇ Z ⁇ 73, it is judged that double materials are stored in the packaging box.
  • the steps of processing image a specifically include:
  • Original image acquisition step obtain a grayscale image of the material placement area in the packaging box, select the entire image area as the target area to be analyzed, and obtain the grayscale image of the entire image area as the original image A, which are image a in Figure 1 and image a in Figure 2.
  • Image reduction step The original image A is reduced in proportion to obtain a reduced image B. That is, the original image A is reduced to one-tenth of the width and one-tenth of the height of the original image A, obtaining a smaller grayscale image B.
  • the original image A is 1020*500 pixels
  • the image B is 102*50 pixels.
  • Image enlargement step enlarge image B to the size of the original image A, and obtain image C of the same size as the original image A. That is, enlarge image B to the size of the original image A, and obtain image C of the same size as the original image A.
  • the original image A is 1020*500 pixels
  • image C is also 1020*500 pixels.
  • Image processing steps Subtract image C from the original image A to obtain image D, and multiply image D by the corresponding coefficient 8 to obtain image E, which are image b in Figure 1 and image b in Figure 2.
  • the grayscale range of the grayscale image is 0 to 255. If the grayscale value of a pixel in image C exceeds 255 after multiplying it by 8, the pixel is set to 255.
  • Confirmation step Calculate the variance of the grayscale value of the target area according to image E to obtain the variance value Z. If the variance value Z is within the range of 49 ⁇ Z ⁇ 73, it is judged that the quantity and type of materials put into the packaging box are correct and it is a qualified product. If the variance value Z is not within the range of 49 ⁇ Z ⁇ 73, it is judged that the quantity and type of materials put into the packaging box are incorrect and it is a unqualified product. The packaging equipment removes it from the packaging line.
  • the variance of the grayscale image of image b in Figure 1 is 44.8119
  • the variance of the grayscale image of image b in Figure 2 is 87.2330.
  • Their variance values Z are not in the range of 49 ⁇ Z ⁇ 73, and they are all unqualified packaged products.
  • the packaging equipment removes them from the packaging line.
  • the variance of the grayscale image of image b in Figure 3 is 63.4496
  • the variance value Z is in the range of 49 ⁇ Z ⁇ 73, which is a qualified packaged product.
  • the cotton swabs in the packaging box are allowed to be placed upside down or upside down, and they are judged as qualified products regardless of whether they are placed upside down or upside down.
  • one side of the cotton swab outer packaging is made of paper material, it is not easy to produce reflections under light, and the other side of the cotton swab outer packaging is made of plastic material, which is easy to produce reflections under light.
  • the cotton swab in the packaging box is placed with the paper surface facing upward
  • FIG5 is placed with the paper surface facing downward and the plastic surface facing upward.
  • the light irradiates the plastic material to produce obvious reflections.
  • the grayscale image in the packaging box is obtained, such as the image a in FIG3 and the image a in FIG5 .
  • the image a is processed by the method 2 of the present invention, and the variance of the grayscale image of the image b in FIG3 is 63.4496, and the variance of the grayscale image of the image b in FIG5 is 58.3440.
  • the steps of processing image a specifically include:
  • Original image acquisition step obtain a grayscale image of the material placement area in the packaging box, select the entire image area as the target area to be analyzed, and obtain a grayscale image of the entire image area as the original image A, which are image a in FIG3 and image a in FIG5 .
  • Image reduction step The original image A of FIG3 is reduced in proportion to obtain a reduced image B. That is, the original image A is reduced to one tenth of the width and one tenth of the height of the original image A to obtain a smaller grayscale image B.
  • the original image A is 1020*500 pixels
  • the image B is 102*50 pixels.
  • Image enlargement step enlarge image B to the size of the original image A, and obtain image C of the same size as the original image A. That is, enlarge image B to the size of the original image, and obtain image C of the same size as the original image.
  • the original image A is 1020*500 pixels
  • image C is also 1020*500 pixels.
  • Image processing steps Subtract image C from the original image A to obtain image D, and multiply image D by the corresponding coefficient 8 to obtain image E, which are image b in Figure 3 and image b in Figure 5.
  • the grayscale range of the grayscale image is 0 to 255. If the grayscale value of a pixel in image C exceeds 255 after multiplying it by 8, the pixel is set to 255.
  • Confirmation step Calculate the variance of the grayscale value of the target area according to image E to obtain the variance value Z. If the variance value Z is greater than 55, it is judged that the cotton swab is placed in the packaging box, which is a qualified product. If the variance value Z is less than 55, it is judged that the cotton swab is not placed in the packaging box, which is a defective product, and the packaging equipment removes it from the packaging line.
  • step 5 image processing step: use the original image A Image D is obtained by subtracting image C, and image D is multiplied by the corresponding coefficient 8 to obtain image E.
  • a fixed grayscale value such as 10 is assigned to the pixels in the F+ area of image E, which are image b in Figure 3 and image b in Figure 5.
  • the grayscale range of the grayscale image is 0 to 255. If the grayscale value of a pixel in image C exceeds 255 after multiplying it by 8, the pixel is set to 255.
  • the variance of the grayscale image of the b image in FIG3 is 63.4496
  • the variance value Z is greater than 55
  • the packaging box has been filled with a cotton swab
  • the variance of the grayscale image of the b image in FIG5 is 58.3440
  • the variance value Z is greater than 55
  • the packaging box has also been filled with a cotton swab. Therefore, by using the method of the present invention, even if the material produces a large reflection under the illumination of the light, it will not affect the correct judgment of the packaging instrument.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Processing (AREA)

Abstract

本发明提供了一种检测包装盒内物料放置情况的方法，包括原始图像获取步骤，缩小图像步骤，放大图像步骤，图像处理步骤和确认步骤等。通过对原图进行缩小和放大处理可快速判断出包装盒内相关物料的存放情况，实现包装线的高度自动化，消除因人工目视检测时的漏检情况。

Description

一种检测包装盒内物料放置情况的方法 技术领域
本发明涉及产品包装技术领域，更具体的说，它涉及一种检测包装盒中物料放置情况的方法。
背景技术
在产品自动化包装生产线上，目前多靠人工目视检测传送带上的包装盒内是否包含相关物料。例如在单人份抗原检测产品自动化包装生产线上，需要向包装盒内投放棉签和合格证，由于生产线产量高，长期依靠人工目视进行检测，眼睛容易疲劳，从而发生漏检，导致产品质量缺陷。虽然包装生产线上会有一道称重工序，称量包装后产品的重量，保证所有物料均完成投放。但棉签和合格证都是轻质的，该称重工序无法检测出棉签和合格证是否投放。
为了提高包装生物线的自动化能力，市面上出现了利用机器视觉***判断包装盒内物料放置情况的设备，而目前市面上已有的设备都是需要放置2套独立的机器视觉***或一个工控机带2个摄像头。采用2套独立的机器视觉***或一个工控机带2个摄像头来检测棉签和合格证有无。其第一套机器视觉***放置在发卡机(合格证喂盒)后，利用模板匹配，根据合格证正面的具体图案特征(例如检测图片上是否存在合格证上的圆)，来检测合格证的有无。第二套机器视觉***放在棉签蜘蛛手(棉签喂盒)后，检测盒内是否有强反光(此时棉签反面朝上，可看到棉签)，或利用模板匹配(此时棉签正面朝上，根据棉签正面的具体图像特征，例如角落长方形的黑标)，来检测棉签的有无。
但是使用2套机器视觉***，硬件成本极高，设备占地面积广，且产品线只能适应单独的一种合格证或棉签，若合格证或棉签的外包装有了较大变化(比如合格证上的圆的直径有较大的变化)，则需要新建模板，以适应新的图案特征，从而导致生产成本升高且生产效率降低。
发明内容
本发明克服了现有技术的不足，提供了一种设计简单、成本低，具有较高的产品适应性的包装方法，本发明的技术方案如下。
本发明提供了一种检测包装盒内物料放置情况的方法，包括以下步骤：
1)原始图像获取步骤：获取包装盒内物料放置区域的灰度图，选定要分析的目标区域，获得该目标区域的灰度图作为原始图像A；
2)缩小图像步骤：将步骤1)中的原始图像A按比例缩小，得到图像B；
3)放大图像步骤：将步骤2)中的图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C；
4)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数或将图像D乘以相应系数再加上一定值，得到图像E；
5)确认步骤：根据图像E计算目标区域的灰度值的方差，得到方差值Z，如果方差值Z在设定的区间范围内，判断包装盒内的物料符合规定；如果方差值Z不在设定的区间范围内，判断包装盒内的物料不符合规定。
进一步的，步骤1)中还包括原始图像A二值化的步骤，将二值化处理后的图像与阈值比较，如果二值化处理后的图像灰度值小于阈值，判断原始图像A没有高亮区域，进入步骤2)；如果二值化处理后的图像中有灰度值大于阈值的部分，则存在高亮部分，需将原始图像A二值化得到高亮部分的区域F进行膨胀得到区域F+，并对F+进行赋值。
进一步的，对区域F+进行赋值选择以下处理方式之一：在步骤1)中将F+区域内像素点赋予固定的灰度值，将该图像作为原始图像A进入步骤2)，或在步骤4)中对图像E中的F+区域内像素点赋予固定的灰度值后，进入步骤5)。
进一步的，所述赋值为0或低亮度值。
进一步的，步骤4)中，D乘以相应系数或将图像D乘以相应系数再加上一定值，若图像D的像素点灰度值经上述计算后超过255，则该像素点取255。
进一步的，步骤4)中所述的相应系数为大于0的数值，所述一定值为正数或负数。
本发明还提供二楼一种包装方法，包括提供一种包装设备，所述包装设备包括开关电源、视觉控制器、剔除装置、传送包装盒的传送带和位于传送带上方的相机，视觉控制器与相机电连接接收相机拍摄的包装盒内部图像，视觉控制器判断包装盒内的物料是否符合规定，若判断包装盒内的物料不符合规定时，视觉控制器输出信号，剔除装置启动，将不合格的包装盒从传送带上剔除，所述判断包装盒内物料是否符合规定的步骤包括：
1)原始图像获取步骤：用包装设备上的相机获取传送带上包装盒内物料放置区域的灰度图，选定要分析的目标区域，获得该目标区域的灰度图作为原始图像A；
2)缩小图像步骤：将步骤1)中的原始图像A按比例缩小，得到图像B；
3)放大图像步骤：将步骤2)中的图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C；
4)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数或将图像D乘以相应系数再加上一定值，得到图像E；
5)确认步骤：根据图像E计算目标区域的灰度值的方差，得到方差值Z，如果方差值Z在设定的区间范围内，判断包装盒内的物料符合规定；如果方差值Z不在设定的区间范围内，判断包装盒内的物料不符合规定。
进一步的，所述包装设备还包括用于感应包装盒进入相机拍摄区域的触发感应器，所述触发感应器与相机电连接。
进一步的，触发感应器上为光电开关。
进一步的，包装设备还包括剔除落口。
进一步的，步骤1)中还包括原始图像A二值化步骤，将二值化处理后的图像与阈值比较，如果二值化处理后的图像灰度值小于阈值，判断原始图像A没有高亮区域，进入步骤2)；如果二值化处理后的图像中有灰度值大于阈值的部分，则存在高亮部分，需将原始图像A二值化得到高亮部分的区域F进行膨胀得到区域F+。可将F+区域内像素点赋予固定的灰度值，将该图像做为原始图像A进入步骤2)。也可在步骤4)中对图像E中的F+区域内像素点赋予固定的灰度值后，再进入步骤5)。在有大面积高亮区域的情况下，对F+区域内像素点赋予固定的灰度值这一步骤可以只选其中一步或均选。
进一步的，步骤4)中所述的相应系数为大于0的数值，所述一定值为正数或负数。
本发明相比现有技术优点在于：本发明通过优化设计，能取代人工作业。且硬件设备只有1套机器视觉***，成本进一步降低，设备的占地面积也大大缩小，有助于开展更多生产线。本发明作业员可实时根据新物料，更改文件内数值，适配新产品，小的局部图案变化不影响视觉***运行，兼容性强。
附图说明
图1为包装盒内没有存放物料的图像。
图2为包装盒内合格证正面朝上放置的图像。
图3为包装盒内放入棉签和合格证的图像，且棉签的纸面朝上放置。
图4为包装盒内合格证正面朝下放置的图像。
图5为包装盒内棉签包装袋的塑料面朝上放置的图像。
图6为本发明判断包装盒物料方式情况的流程示意图。
图7为本发明的设备配置电路示意图。
图8为本发明包装设备的示意图。
图9为本发明包装设备另一个角度的示意图。
具体实施方式
下面结合附图和具体实施例对本发明的包装方法进行详细描述。
如图7、图8和图9所示的包装设备100，包括设备的开关电源、传送包装盒的传送带101，包装盒200相互之间间隔一定距离并依次排放在传送带101上，相机102和相机光源103分别安装在支架106上。相机102和相机光源103安装在传送带101的正上方，相机光源103位于相机和传送带101之间。相机102直接安装在支架106上，相机光源103通过光源支架107与支架106连接。相机光源103采用的是圆环灯，在光源支架107上开设了拍摄通孔108，相机102通过拍摄通孔108和圆环灯的中心无灯区域，对准包装盒内部进行拍摄。触发感应器104安装在传送带101的安装支架109上，触发感应器104(图7中SQ1)，当包装盒输送至触发感应器处，触发感应器感应到说明包装盒已进入相机拍摄区，触发信号(光电开关可设置为上升沿触发，下降沿触发，接通触发等情况)通过电路直接触发相机拍照，视觉控制器定时扫描中收到相机拍摄出这个包装盒的内部照片。视觉控制器根据图6所示的判断包装盒投料情况流程图所示的方法，判断出包装盒内的投料是否符合规定，若判断符合规定，则该合格的包装盒继续沿传送带前行，若判断投料不符合规定，例如包装盒缺少物料或物料方向放反等情况，则视觉控制器的输出点(图7中视觉控制器的1所示)有输出，从而接通中间继电器KA1的线圈，KA1的常开触点闭合，剔除电磁阀YV1导通，包装设备的剔除装置将该不合格的包装盒推入包装盒剔除落口105内。剔除装置和包装盒剔除落口位于触发感应器的下游。
相机光源的安装位置也可以安装在包装设备的其他位置；或者在确保拍照所需亮度的情况下，也可以不安装相机光源；或者相机光源直接装配在相机上。
触发感应器的安装位置也可以安装在包装设备的其他位置，并可以采用能判断出包装盒已进入拍摄区的其他元件。触发感应器也可以不安装，相机连续拍照，当视觉控制器判断投料符不合规定时，剔除装置启动，将不合格的包装盒导出。
剔除装置采用剔除电磁阀，剔除启动时，剔除装置的气管会接通，通过喷气的方式将不合格的包装盒吹出传送带。剔除装置还可以采用其他方式。
下面结合图6具体说明判断包装盒内物料是否符合规定的方法。
方法一
以下是本发明判断包装盒内物料放置是否符合规定的一种包装方法，具体包括以下步骤：
1)原始图像获取步骤：获取包装盒内物料放置区域的灰度图，选定要分析的目标区域，获得该目标区域的灰度图作为原始图像A。所述目标区域可以是包装盒内全图区域或全图区域内的某部分区域。
2)缩小图像步骤：将原始图像A按比例缩小，得到缩小的图像B。
3)放大图像步骤：将图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C。
4)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数或将图像D乘以相应系数再加上一定值，得到图像E。所述相应系数为大于0的数值，所述一定值可以是正数也可以是负数。相应系数和一定值可以根据测试确定。
5)确认步骤：根据图像E计算目标区域的灰度值的方差，得到方差值Z。如果方差值Z在设定的区间范围内，则包装盒内物料放置符合规定，确认为合格品；如果方差值Z不在设定的区间范围内，则包装盒内物料放置不符合固定，确认为不合格品。
方法二
以下是本发明判断包装盒内物料放置是否符合规定的一种包装方法，该方法是针对图片中存在高亮部分时的一种处理方法，相比于方法一的不同点是在获得原始图像A后需判断是否存在高亮区域，并对高亮区域进行处理。具体包括以下步骤：
1)原始图像获取步骤：获取包装盒内物料放置区域的灰度图，选定要分析的目标区域，获得该目标区域的灰度图作为原始图像A。所述目标区域可以是包装盒内全图区域或全图区域内的某部分区域。
2)高亮区域的判断和处理步骤：原始图像A二值化，将二值化处理后的原图与阈值进行比较，如果二值化处理后的原图的灰度值小于阈值，判断原图没有高亮区域；如果二值化处理后的原始图像A中有灰度值大于阈值的部分，这部分即为高亮部分，则需将原始图像A二值化得到高亮部分的区域F进行膨胀得到区域F+。
3)缩小图像步骤：若经步骤2)判断出原始图像A没有高亮区域，则将原始图像A按比例缩小，得到缩小的图像B；若经步骤2)判断出原始图像A存在高亮区域，则对高亮部分的区域进行膨胀处理后的图像按比例缩小，得到缩小的图像B。
4)放大图像步骤：将图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C。
5)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数或将图像D乘以相应系数再加上一定值，得到图像E。所述相应系数为大于0的数值，所述一定值可以是正数也可以是负数。相应系数和一定值可以根据测试确定。
6)确认步骤：根据图像E计算目标区域的灰度值的方差，得到方差值Z。如果方差值Z在设定的区间范围内，则包装盒内物料放置符合规定，确认为合格品；如果方差值Z不在设定的区间范围内，则包装盒内物料放置不符合固定，确认为不合格品。
方法二中还包括对区域F+进行赋值的步骤。赋值步骤可以在步骤2)中进行，即高亮部分的区域F进行膨胀得到区域F+，将原始图像A的区域F+灰度值赋值为0或低亮度值。赋值步骤也可以改在步骤5)中进行，即将图像E中的区域F+灰度值赋值为0或低亮度值。
实施例1不同处理方式的灰度值的方差比较
图1是包装盒内没有存放棉签和合格证这两种物料的图像，其中图1中的a图像是未经本发明方法处理，通过拍摄包装盒内物料放置区域后直接获得灰度图图像，其灰度值方差为22.3106，图1中的b图像是经本发明方法一处理后的图像，其灰度值方差为44.8119。
图2是包装盒内放入合格证的图像，其中图2中的a图像是未经本发明方法处理，通过拍摄包装盒内物料放置区域后直接获得灰度图图像，其灰度值方差为31.3447，图2中的b图像是经本发明方法一处理后的图像，其灰度值方差为87.2330。
图3是包装盒内放入棉签和合格证的图像，其中图3中的a图像是未经本发明方法处理，通过拍摄包装盒内物料放置区域后直接获得灰度图图像，其灰度值方差为35.8878，图3中的b图像是经本发明方法二处理后的图像，其灰度值方差为63.4496。
从未经本发明方法处理，而是通过拍摄包装盒内物料放置区域后直接获得灰度图图像分析结果看，仅放了合格证的图像灰度值方差与未放合格证的图像灰度值方差之间的差值仅为9.0341(31.3447-22.3106)，棉签和合格证都放了的图像灰度值方差与仅放了合格证的图像灰度值方差之间的差值仅为4.5431(35.8878-31.3447)。这三种情况的灰度值方差值相差很小，分析软件无法通过方差值分析出包装盒内是否放了相关物料，例如是否放了合格证、棉签等，或者由于图像测量偏差也可能会造成方差的细微波动，如果灰度值方差值相差很小，包装设备就无法判断方差值是因测量偏差，还是的确没有放入正确的物料。
从经过本发明方法处理的图像分析结果看，仅放了合格证的图像灰度值方差与未放合格 证的图像灰度值方差之间的差值为42.4211(87.2330-44.8119)，仅放了合格证的图像灰度值方差与棉签和合格证都放了的图像灰度值方差之间的差值为23.7834(87.2330-63.4496)。这三种情况的灰度值方差值相差很大，图像的测量偏差对图像灰度值方差不会产生较大影响，因此包装设备能够准确判断出包装盒内是否放入了正确的物料。
因此，本发明的包装方法能有效且准确的判断出包装盒内是否放入了正确的物料。
实施例2判断包装盒内物料的放置方向是否正确
如图2是将合格证的正面朝上的放入包装盒内，图4是将合格证正面朝下的放入包装盒内。合格证的正面是指带有文字的这一面，合格证的反面是指没有文字的这一面。放置完成后，获取包装盒内的灰度图，如图2中的a图像和图4中的a图像。然后采用本发明的方法一分别对a图像进行图像处理，获得图2的b图像的灰度图的方差为87.2330，获得图4的b图像的灰度图的方差为42.325。本实施例中设定合格证正面朝上时，灰度图的方差值Z要大于设定值66。图2的b图像的灰度图的方差为87.2330，其大于设定值66，包装机判断出包装盒内的产品合格证是正面朝上放置，为正确放置，图4的b图像的灰度图的方差为42.325，其小于设定值66，包装机判断出包装盒内的产品合格证是正面朝下放置，为错误放置，需要从包装线上剔除。
实施例3判断包装盒内是否有物料
如图1是包装盒内没有放入物料，即没有放入产品合格证，图2是包装盒内放入了物料，即放入了产品合格证。放置完成后，获取包装盒内的灰度图，如图1中的a图像和图2中的a图像。然后采用本发明的方法一分别对a图像进行图像处理，获得图1的b图像的灰度图的方差为44.8119，获得图2的b图像的灰度图的方差为87.2330。本实施例中设定合格证正面朝上时，灰度图的方差值Z要大于设定值60。图1的b图像的灰度图的方差为44.8119，其小于设定值60，包装机判断出包装盒内没有放入物料，是空盒，为错误放置，需要从包装线上剔除，图2的b图像的灰度图的方差为87.2330，其大于设定值60，包装机判断出包装盒内的已经放入物料，为正确放置。
实施例4判断包装盒中物料放入数量和种类是否正确
如图1是包装盒内没有放入物料，即没有放入产品合格证，图2是包装盒内放入了单一物料，即放入了合格证。图3是包装盒内放入了双物料，即放入了合格证和棉签。放置完成 后，获取包装盒内的灰度图，如图1中的a图像、图2中的a图像和图3中的a图像。然后采用本发明的方法一分别对a图像进行图像处理，获得图1的b图像的灰度图的方差为44.8119，获得图2的b图像的灰度图的方差为87.2330，获得图3的b图像的灰度图的方差为63.4496。本实施例设定方差值Z在49<Z<73内时，判断包装盒内存放了双物料。
本实施例中对a图像的处理步骤具体包括：
1)原始图像获取步骤：获取包装盒内物料放置区域的灰度图，选定取全图区域为要分析的目标区域，获得全图区域的灰度图作为原始图像A，分别是图1中的a图像和图2中a图像。
2)缩小图像步骤：将原始图像A按比例缩小，得到缩小的图像B。即缩小原始图像A到原始图像A大小的十分之一的宽度和十分之一的高度，得到更小尺寸的灰度图像B。在本例中原始图像A为1020*500像素，图像B则为102*50像素。
3)放大图像步骤：将图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C。即放大图像B到原始图像A大小，得到和原始图像A同样大小的图像C。在本例中原始图像A为1020*500像素，图像C也为1020*500像素。
4)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数8，得到图像E，分别是图1中的b图像和图2中的b图像。灰度图的灰度范围0～255，若图像C的某像素点灰度值乘以8后超过255，则该像素点取255。
5)确认步骤：根据图像E计算目标区域计算灰度值的方差，得到方差值Z。如果方差值Z在49<Z<73范围内，则判断包装盒的物料放入数量和种类都正确，为合格品，如果方差值Z不在49<Z<73范围内，则判断包装盒的物料放入数量和种类不正确，为不合格品，包装设备将其从包装流水线上剔除。
在本例中图1的b图像的灰度图的方差为44.8119，图2的b图像的灰度图的方差为87.2330，它们的方差值Z都不在49<Z<73范围内，均为不合格的包装产品，包装设备将它们从包装流水线上剔除，图3的b图像的灰度图的方差为63.4496，方差值Z在49<Z<73范围内，为合格的包装产品。
实施例5包装盒中物料放置方向对判断结果的影响
在本例中，包装盒内的棉签正反放置都是允许的，无论正放还是反放都判断为合格品。但由于棉签外包装的一个面是纸质材料，在灯光照射下不易产生反光，棉签外包装的另一面为塑料材质，在灯光照射下容易产生反光。
如图3是包装盒内的棉签为纸面朝上的放置方式，图5是包装盒内的棉签为纸面朝下和塑料材质面朝上的放置方式，灯光照射在塑料材质上产生明显的反光。放置完成后，获取包装盒内的灰度图，如图3中的a图像和图5中的a图像。然后采用本发明的方法二分别对a图像进行图像处理，获得图3的b图像的灰度图的方差为63.4496，获得图5的b图像的灰度图的方差为58.3440。
本实施例中对a图像的处理步骤具体包括：
1)原始图像获取步骤：获取包装盒内物料放置区域的灰度图，选定取全图区域为要分析的目标区域，获得全图区域的灰度图作为原始图像A，分别是图3中的a图像和图5中的a图像。
2)高亮区域的判断和处理步骤：原始图像A二值化，设定的阈值为180。将二值化处理后的原图与阈值进行比较。如果二值化处理后的原始图像A的灰度值小于阈值180，判断原始图像A没有高亮区域，二值化后的图像为全黑，图3就是这类情况，无需进行膨胀。如果二值化处理后的原始图像A中有灰度值大于阈值180的部分，这大于阈值180部分即为高亮部分，图5就是这类情况，则需将图5原始图像A二值化得到高亮部分的区域F进行膨胀得到区域F+，本例中区域F进行10个像素核大小的圆形膨胀得到区域F+。将F+区域内像素点赋予固定的灰度值如10，将该图像做为原始图像A进入步骤3)。
3)缩小图像步骤：将图3的原始图像A的图像按比例缩小，得到缩小的图像B。即缩小原始图像A到原始图像A大小的十分之一的宽度和十分之一的高度，得到更小尺寸的灰度图像B。在本例中原始图像A为1020*500像素，图像B则为102*50像素。
4)放大图像步骤：将图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C。即放大图像B到原图大小，得到和原图同样大小的图像C。在本例中原始图像A为1020*500像素，图像C也为1020*500像素。
5)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数8，得到图像E，分别是图3中的b图像和图5中的b图像。灰度图的灰度范围0～255，若图像C的某像素点灰度值乘以8后超过255，则该像素点取255。
6)确认步骤：根据图像E计算目标区域计算灰度值的方差，得到方差值Z。如果方差值Z大于55，则判断包装盒的放入了棉签，为合格品，如果方差值Z小于55，则判断包装盒内未放入棉签，为不合格品，包装设备将其从包装流水线上剔除。
另一个对a图像的处理步骤中，有关上述步骤2)中的“将F+区域内像素点赋予固定的灰度值如10”这一步也可以选择在步骤5)中进行，即步骤5)图像处理步骤：用原始图像A 减去图像C后得到图像D，将图像D乘以相应系数8，得到图像E，对图像E中的F+区域内像素点赋予固定的灰度值如10，分别是图3中的b图像和图5中的b图像。灰度图的灰度范围0～255，若图像C的某像素点灰度值乘以8后超过255，则该像素点取255。
在本例中图3的b图像的灰度图的方差为63.4496，方差值Z大于55，包装盒内已经装有棉签，图5的b图像的灰度图的方差为58.3440，方差值Z大于55，包装盒内同样已经装有棉签。因此采用本发明的方法，即使物料在灯光照射下产生大的反光，也不会影响包装仪做出正确的判断。

Claims (6)

  1. 一种检测包装盒内物料放置情况的方法，其特征在于，包括以下步骤：
    1)原始图像获取步骤：获取包装盒内物料放置区域的灰度图，选定要分析的目标区域，获得该目标区域的灰度图作为原始图像A；
    2)缩小图像步骤：将步骤1)中的原始图像A按比例缩小，得到图像B；
    3)放大图像步骤：将步骤2)中的图像B放大至原始图像A的大小，得到和原始图像A同样尺寸的图像C；
    4)图像处理步骤：用原始图像A减去图像C后得到图像D，将图像D乘以相应系数或将图像D乘以相应系数再加上一定值，得到图像E；
    5)确认步骤：根据图像E计算目标区域的灰度值的方差，得到方差值Z，如果方差值Z在设定的区间范围内，判断包装盒内的物料符合规定；如果方差值Z不在设定的区间范围内，判断包装盒内的物料不符合规定。
  2. 根据权利要求1所述的方法，其特征在于，步骤1)中还包括原始图像A二值化的步骤，将二值化处理后的图像与阈值比较，如果二值化处理后的图像灰度值小于阈值，判断原始图像A没有高亮区域，进入步骤2)；如果二值化处理后的图像中有灰度值大于阈值的部分，则存在高亮部分，需将原始图像A二值化得到高亮部分的区域F进行膨胀得到区域F+，并对F+进行赋值。
  3. 根据权利要求2所述的方法，其特征在于，对区域F+进行赋值选择以下处理方式之一：在步骤1)中将F+区域内像素点赋予固定的灰度值，将该图像作为原始图像A进入步骤2)，或在步骤4)中对图像E中的F+区域内像素点赋予固定的灰度值后，进入步骤5)。
  4. 根据权利要求3所述的方法，其特征在于，所述赋值为0或低亮度值。
  5. 根据权利要求1所述的方法，其特征在于，步骤4)中，D乘以相应系数或将图像D乘以相应系数再加上一定值，若图像D的像素点灰度值经上述计算后超过255，则该像素点取255。
  6. 根据权利要求1所述的方法，其特征在于，步骤4)中所述的相应系数为大于0的数值，所述一定值为正数或负数。
PCT/CN2023/132500 2022-11-21 2023-11-20 一种检测包装盒内物料放置情况的方法 WO2024109669A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211469753.8 2022-11-21
CN202211469753.8A CN115908328A (zh) 2022-11-21 2022-11-21 一种检测包装盒内物料放置情况的方法

Publications (1)

Publication Number Publication Date
WO2024109669A1 true WO2024109669A1 (zh) 2024-05-30

Family

ID=86484420

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/132500 WO2024109669A1 (zh) 2022-11-21 2023-11-20 一种检测包装盒内物料放置情况的方法

Country Status (2)

Country Link
CN (1) CN115908328A (zh)
WO (1) WO2024109669A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115908328A (zh) * 2022-11-21 2023-04-04 利多(香港)有限公司 一种检测包装盒内物料放置情况的方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104020181A (zh) * 2014-05-26 2014-09-03 中国电子科技集团公司第三十八研究所 用无损检验装置对集成电路物料进行检测的方法
US20170287252A1 (en) * 2016-04-03 2017-10-05 Harshal Dwarkanath Laddha Counterfeit Document Detection System and Method
WO2018107525A1 (zh) * 2016-12-12 2018-06-21 广东省智能制造研究所 一种电动注塑机模具检测方法
CN110866911A (zh) * 2019-11-14 2020-03-06 华雁智能科技(集团)股份有限公司 表盘缺陷检测方法、装置、图像处理设备及存储介质
CN111539980A (zh) * 2020-04-27 2020-08-14 航天科工微电子***研究院有限公司 一种基于可见光的多目标追踪方法
CN115908328A (zh) * 2022-11-21 2023-04-04 利多(香港)有限公司 一种检测包装盒内物料放置情况的方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104020181A (zh) * 2014-05-26 2014-09-03 中国电子科技集团公司第三十八研究所 用无损检验装置对集成电路物料进行检测的方法
US20170287252A1 (en) * 2016-04-03 2017-10-05 Harshal Dwarkanath Laddha Counterfeit Document Detection System and Method
WO2018107525A1 (zh) * 2016-12-12 2018-06-21 广东省智能制造研究所 一种电动注塑机模具检测方法
CN110866911A (zh) * 2019-11-14 2020-03-06 华雁智能科技(集团)股份有限公司 表盘缺陷检测方法、装置、图像处理设备及存储介质
CN111539980A (zh) * 2020-04-27 2020-08-14 航天科工微电子***研究院有限公司 一种基于可见光的多目标追踪方法
CN115908328A (zh) * 2022-11-21 2023-04-04 利多(香港)有限公司 一种检测包装盒内物料放置情况的方法

Also Published As

Publication number Publication date
CN115908328A (zh) 2023-04-04

Similar Documents

Publication Publication Date Title
WO2024109669A1 (zh) 一种检测包装盒内物料放置情况的方法
CN108627457B (zh) 自动光学检测***及其操作方法
CN103072718B (zh) 用于食品包装生产线的机器视觉检测***及方法
KR102168724B1 (ko) 이미지 검사를 이용한 이상 판별 방법 및 장치
CN103364400A (zh) 一种多工位智能融合的瓶装啤酒在线检测装置
CN110246122A (zh) 基于机器视觉的小型轴承质量检测方法、装置及***
WO2009009180A2 (en) Method and system for inspecting manufactured parts and sorting the inspected parts
JP2013040915A (ja) ガラス基板の表面不良検査装置および検査方法
KR101779974B1 (ko) 글래스 크랙 검사 시스템
CN105784719A (zh) 能够进行物品分类的机器视觉检验终端机
WO2006088150A1 (ja) ガラス基板検査装置および検査方法
CN107525812A (zh) 一种包装袋喷码缺陷检测方法
JP2000180382A (ja) 外観検査装置
CN109701890A (zh) 磁瓦表面缺陷检测与分拣方法
KR20190020589A (ko) 다중 방식의 제품 검사시스템
CN113015901A (zh) 玻璃板的制造方法以及玻璃板的制造装置
TWI753424B (zh) 外觀檢查管理系統、外觀檢查管理裝置、外觀檢查管理方法以及程式
CN113640303A (zh) 一种用于笔记本电脑的表面瑕疵检验设备及其检测方法
KR100948795B1 (ko) 주화검사장치
CN203385393U (zh) 全自动光学检验机
JP2004191074A (ja) タブレットの密度推定方法及び検査装置
JP2000081396A (ja) 巻取ロール側面の品質自動判定方法及び装置
JPH04121647A (ja) 金属缶端部の外観検査方法及び装置
TWI723548B (zh) 邊緣檢查裝置
JPH08136467A (ja) 欠陥検査方法およびその装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23893759

Country of ref document: EP

Kind code of ref document: A1