Food packaging visual inspection device and method in food processing

Abstract

The invention discloses a food packaging visual detection device and method in food processing. The detection device comprises a loading manipulator, a turntable feeding mechanism, a visual detection mechanism, an unloading manipulator, a non-defective product conveying belt, a defective product conveying belt and a temporary storage box, the rotary disc feeding mechanism comprises a supporting rotary table, a mounting disc is arranged on the supporting rotary table, a plurality of containing assemblies are arranged on the upper surface of the mounting disc, and a vibration motor is arranged in the middle of each containing assembly. The vibration motor can enable packaging bags placed on the placing assembly to vibrate, for some packaging bags which are prone to wrinkling or unevenness, vibration can enable the surfaces of the packaging bags to be smoother, the situation that the detection result of the visual detection mechanism is affected by wrinkling or unevenness of the surfaces of the packaging bags is avoided, the detection accuracy is improved, and misjudgment is reduced; according to the detection method, the detection process is divided into two stages, the wrinkle coefficient is judged firstly, then surface feature analysis is carried out, and the detection process is more accurate and efficient through the staged detection mode.

Description

Technical Field

[0001] The present invention belongs to the technical field of visual inspection of food packaging, and particularly relates to a visual inspection device and method for food packaging during food processing. Background Art

[0002] In the food processing industry, the quality of food packaging is directly related to food safety and market image. However, various defects may occur during the production process of food packaging, such as scratches, breakages, stains, printing errors, etc. Traditional manual inspection methods are inefficient, prone to missed inspections and misjudgments, and have a high labor intensity. Therefore, visual inspection devices are usually used to replace traditional manual inspections to improve inspection efficiency and reduce the labor intensity of operators.

[0003] However, if the packaging bag is not fully flattened during transportation in the existing visual inspection device, creases will occur, and the creases on the surface of the packaging bag are likely to cause misjudgments by the visual inspection device, and excessive misjudgment phenomena are likely to affect the effective utilization rate of the equipment.

[0004] Therefore, the present application provides a visual inspection device and method for food packaging during food processing to reduce the misjudgment rate during inspection and improve the reliability of inspection. Summary of the Invention

[0005] The present invention provides a visual inspection device and method for food packaging during food processing, aiming to solve the problems pointed out in the above background art.

[0006] To solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A visual inspection device for food packaging during food processing, comprising: a loading manipulator, a turntable feeding mechanism, a visual inspection mechanism, an unloading manipulator, a good product conveyor belt, a defective product conveyor belt, and a buffer box;

[0008] The feeding manipulator is used to transfer the packaging bags to be tested to the turntable feeding mechanism, and the turntable feeding mechanism is used to transfer the packaging bags to be tested to the detection area of the vision detection mechanism; the vision detection mechanism is used to perform vision detection on the packaging bags and distinguish the packaging bags into qualified packaging bags and defective packaging bags according to the detection results; specifically, the vision detection mechanism is used to collect the image information of the packaging bags, and then preprocess the collected image information, and the preprocessing includes filtering, noise reduction and enhancement operations; and use the edge detection algorithm to extract the edge information of the packaging bags; calculate the number of pixels in the area enclosed by the edges; then convert the pixel area into the actual area according to the conversion relationship between the pre-calibrated pixels and the actual size, so as to obtain the image area of the packaging bag; and calculate the ratio of the image area of the packaging bag to the calibrated area of the packaging bag, and use it as the wrinkle coefficient, and the calibrated area of the packaging bag is the actual area when the packaging bag is completely flattened; finally, judge whether the wrinkle coefficient exceeds the preset range, if so, judge it as a suspicious packaging bag, otherwise analyze the surface features of the packaging bag image, and calibrate the packaging bag as a qualified packaging bag or a defective packaging bag according to the surface feature analysis results;

[0009] The discharging manipulator transfers the packaging bags to the qualified conveyor belt or the defective conveyor belt or the buffer box according to the detection results of the vision detection mechanism.

[0010] Further, the turntable feeding mechanism includes: a support turntable, on which an installation disk is horizontally arranged coaxially, and a plurality of groups of placing components are arranged on the upper surface of the installation disk, and a vibration motor is arranged in the middle of the placing components; the placing component includes: a mounting plate, the mounting plate is fixedly connected to the installation disk through fixing bolts, a vibration motor is arranged in the middle of the mounting plate, a placing disk is abutted against the upper part of the vibration motor, four corners of the lower surface of the placing disk are vertically provided with connecting rods, mounting holes matched with the connecting rods are opened on the upper surface of the mounting plate, and a pin hole communicated with the mounting hole is opened on the side wall of the mounting plate, and a pin for fixing the connecting rod is threadedly connected in the pin hole.

[0011] Further, an installation groove is opened in the middle of the upper surface of the mounting plate, a first shock-absorbing layer is arranged on the circumferential side wall of the installation groove, and the vibration motor is in interference fit with the first shock-absorbing layer; a second shock-absorbing layer is directly arranged between the upper surface of the vibration motor and the placing disk.

[0012] Further, the vision detection mechanism includes: a vertical moving component, a camera mounting arm is arranged on the vertical moving component, a camera mounting plate is arranged at the end of the camera mounting arm, a light strip disk is arranged on the lower surface of the camera mounting plate, a detection camera and a plurality of annular light strips are arranged on the lower surface of the light strip disk, the detection camera is located in the middle of the light strip disk, and a plurality of the annular light strips are concentrically distributed with the axis of the detection camera as the center.

[0013] Further, a light-shielding cover is provided on the lower surface of the camera mounting plate. The light-shielding cover and the camera mounting plate cooperate to form a sealed space with an open lower end. A flange is provided at the edge of the upper surface of the placing tray. The shape of the flange is adapted to that of the light-shielding cover and is spaced from the light-shielding cover.

[0014] Further, the structure of the loading manipulator is the same as that of the unloading manipulator. The loading manipulator includes: a mounting seat, a first turntable fixedly connected to the mounting seat, a rotating part of the first turntable fixedly connected to one end of a first extension arm, the other end of the first extension arm fixedly connected to a second turntable, a rotating part of the second turntable fixedly connected to one end of a second extension arm, the other end of the second extension arm fixedly connected to a third turntable, a vertically arranged lifting rod fixedly connected to the working end of the third turntable, and a suction plate fixedly connected to the lower end of the lifting rod.

[0015] Further, the shape of the suction plate is adapted to the shape of the inner side of the flange of the placing tray.

[0016] Further, the present application also provides a visual inspection method for food packaging in food processing. The inspection method uses the inspection device as described above and includes the following steps:

[0017] The loading manipulator transports the packaging bag to the placing tray of the turntable feeding mechanism.

[0018] The supporting turntable of the turntable feeding mechanism rotates to move the placing tray with the packaging bag placed thereon to the inspection area of the visual inspection mechanism. During the movement, the vibration motor works to flatten the packaging bag on the placing tray.

[0019] After the placing tray moves to the inspection area of the visual inspection mechanism and the placing tray completely falls within the vertical projection of the light-shielding cover, the inspection camera acquires the image information of the packaging bag and analyzes and processes the image information to classify the packaging bag into a good-quality packaging bag, a defective packaging bag, and a suspicious packaging bag.

[0020] After the inspection is completed, the supporting turntable rotates to move the placing tray out of the inspection area of the visual inspection mechanism. The unloading manipulator transports the packaging bag to the good-quality conveyor belt, the defective conveyor belt, or the buffer box according to the inspection result.

[0021] Further, the analysis and processing of the image information include:

[0022] Preprocessing the acquired image information, where the preprocessing includes filtering, noise reduction, and enhancement operations.

[0023] Using an edge detection algorithm to extract the edge information of the packaging bag.

[0024] Calculate the number of pixels within the area enclosed by the calculated edge;

[0025] Convert the pixel area to the actual area according to the pre-calibrated conversion relationship between pixels and the actual size, thereby obtaining the image area of the packaging bag;

[0026] Calculate the ratio of the image area of the packaging bag to the calibrated area of the packaging bag, and use it as the wrinkle coefficient. The calibrated area of the packaging bag is the actual area when the packaging bag is completely flattened;

[0027] When the wrinkle coefficient exceeds the preset range, mark the packaging bag as a suspicious packaging bag and do not perform surface feature analysis;

[0028] When the wrinkle coefficient is within the preset range, perform surface feature analysis on the packaging bag image. The surface feature analysis includes surface defect detection and printing quality detection;

[0029] Mark the packaging bag as a good-quality packaging bag or a bad-quality packaging bag according to the surface feature analysis results.

[0030] Compared with the prior art, the present invention has the following technical effects:

[0031] 1. In the food packaging visual inspection device in food processing described in the present invention, a vibration motor is provided in the middle of the placement component. The vibration motor can cause the packaging bag placed on the placement component to vibrate. For some packaging bags that are prone to wrinkling or unevenness, the vibration can make the surface of the packaging bag smoother, avoiding the influence of wrinkles or unevenness on the surface of the packaging bag on the detection results of the visual inspection mechanism, improving the accuracy of detection, and reducing misjudgment.

[0032] 2. The food packaging visual inspection method in food processing described in the present invention divides the detection process into two stages: first, judge the wrinkle coefficient, and then perform surface feature analysis. This staged detection method makes the detection process clearer and more efficient. For suspicious products with more wrinkles, directly mark and classify them without performing subsequent complex surface feature detection; for packaging bags with a wrinkle coefficient within the preset range, perform detailed surface feature analysis. This can focus on accurately detecting packaging bags that may be qualified, improve the overall efficiency of the detection process, reduce misjudgment caused by excessive wrinkles on the surface of the packaging bag, improve the yield rate, and enhance the reliability of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is the overall layout top view of the food packaging visual inspection device in food processing described in the present invention;

[0034] Figure 2 is the schematic diagram of the cooperation between the turntable feeding mechanism and the visual inspection mechanism of the food packaging visual inspection device in food processing described in the present invention;

[0035] Figure 3 It is an axonometric view of the placement component of a food packaging visual inspection device in food processing according to the present invention;

[0036] Figure 4 It is a front view of the placement component of a food packaging visual inspection device in food processing according to the present invention;

[0037] Figure 5 It is an internal schematic diagram of the visual inspection mechanism of a food packaging visual inspection device in food processing according to the present invention;

[0038] Figure 6 It is a schematic structural diagram of the loading manipulator of a food packaging visual inspection device in food processing according to the present invention.

[0039] In the figure:

[0040] 1. Loading manipulator; 101. Mounting base; 102. First turntable; 103. First extension arm; 104. Second turntable; 105. Second extension arm; 106. Third turntable; 107. Suction plate;

[0041] 2. Rotary table feeding mechanism; 201. Support turntable; 202. Mounting disc;

[0042] 203. Placement component; 2031. Mounting plate; 2032. Vibration motor; 2033. Placement tray; 2034. Connecting rod; 2035. Pin hole; 2036. First shock-absorbing layer; 2037. Fixed bolt; 2038. Second shock-absorbing layer;

[0043] 3. Visual inspection mechanism; 301. Vertical movement component; 302. Camera mounting arm; 303. Camera mounting plate; 304. Light strip disc; 305. Detection camera; 306. Ring-shaped light strip; 307. Light-shielding cover;

[0044] 4. Unloading manipulator; 5. Good product conveyor belt; 6. Defective product conveyor belt; 7. Buffer box; 8. Negative pressure conveyor belt; 9. Feeding box. Detailed implementation manners

[0045] To make the objectives, technical solutions and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention in conjunction with specific embodiments of the present application and with reference to the accompanying drawings.

[0046] As Figures 1-4 shown, a food packaging visual inspection device in food processing includes: a loading manipulator 1, a rotary table feeding mechanism 2, a visual inspection mechanism 3, an unloading manipulator 4, a good product conveyor belt 5, a defective product conveyor belt 6, and a buffer box 7;

[0047] The feeding manipulator 1 is used to transfer the packaging bag to be tested to the turntable feeding mechanism 2, and the turntable feeding mechanism 2 is used to transfer the packaging bag to be tested to the detection area of the vision detection mechanism 3; the vision detection mechanism 3 is used to perform vision detection on the packaging bag and classify the packaging bag into a qualified packaging bag and a defective packaging bag according to the detection result; specifically, the vision detection mechanism 3 is used to collect the image information of the packaging bag, and then preprocess the collected image information, and the preprocessing includes filtering, noise reduction and enhancement operations; and use the edge detection algorithm to extract the edge information of the packaging bag; calculate the number of pixels in the area enclosed by the edge; then convert the pixel area into the actual area according to the conversion relationship between the pre-calibrated pixels and the actual size, so as to obtain the image area of the packaging bag; and calculate the ratio of the image area of the packaging bag to the calibrated area of the packaging bag, and use it as the wrinkle coefficient, and the calibrated area of the packaging bag is the actual area when the packaging bag is completely flattened; finally, judge whether the wrinkle coefficient exceeds the preset range, if so, judge it as a suspicious packaging bag, otherwise analyze the surface characteristics of the packaging bag image, and calibrate the packaging bag as a qualified packaging bag or a defective packaging bag according to the surface characteristic analysis result; the discharging manipulator 4 transfers the packaging bag to the qualified conveyor belt 5 or the defective conveyor belt 6 or the buffer box 7 according to the detection result of the vision detection mechanism 3;

[0048] The turntable feeding mechanism 2 includes: a support turntable 201, on which an installation disk 202 is horizontally arranged coaxially, and a plurality of groups of placement components 203 are arranged on the upper surface of the installation disk 202, and a vibration motor 2032 is arranged in the middle of the placement component 203.

[0049] By arranging a vibration motor 2032 in the middle of the placement component 203, the vibration motor 2032 can make the packaging bag placed on the placement component vibrate. For the packaging bag with wrinkles or unevenness during the transfer process, the vibration can make the surface of the packaging bag smoother, avoid affecting the detection result of the vision detection mechanism 3 due to the wrinkles or unevenness on the surface of the packaging bag, improve the detection accuracy and reduce misjudgment.

[0050] Such as Figures 3-4As shown, the placement component 203 includes: a mounting plate 2031, the mounting plate 2031 is fixedly connected to the mounting disc 202 by fixing bolts 2037, a vibration motor 2032 is arranged in the middle of the mounting plate 2031, a placement disc 2033 is abutted against the upper part of the vibration motor 2032, connecting rods 2034 are vertically arranged at the four corners of the lower surface of the placement disc 2033, mounting holes matching with the connecting rods 2034 are formed on the upper surface of the mounting plate 2031, and pin holes 2035 communicating with the mounting holes are formed on the side wall of the mounting plate 2031. A pin for fixing the connecting rod 2034 is threadedly connected in the pin hole 2035. Each component of the placement component 203 is detachably connected, which is convenient to replace corresponding-sized parts according to the specifications of the packaging bag products, and is also convenient to replace damaged parts in time.

[0051] As Figures 3-4 shown, an installation groove is formed in the middle of the upper surface of the mounting plate 2031, a first shock-absorbing layer 2036 is arranged on the circumferential side wall of the installation groove, and the vibration motor 2032 is in interference fit with the first shock-absorbing layer 2036; a second shock-absorbing layer 2038 is directly arranged between the upper surface of the vibration motor 2032 and the placement disc 2033. Through the arrangement of the shock-absorbing layer, it can be avoided that the vibration motor 2032 is in direct contact with the placement disc 2033 and the mounting plate 2031, reducing the risk of component damage.

[0052] As Figure 2 and Figure 5 shown, the vision detection mechanism 3 includes: a vertical movement component 301, a camera mounting arm 302 is arranged on the vertical movement component 301, a camera mounting plate 303 is arranged at the end of the camera mounting arm 302, a light strip disc 304 is arranged on the lower surface of the camera mounting plate 303, a detection camera 305 and a plurality of annular light strips 306 are arranged on the lower surface of the light strip disc 304, the detection camera 305 is located in the middle of the light strip disc 304, and the plurality of annular light strips 306 are concentrically distributed with the axis of the detection camera 305 as the center.

[0053] Through the arrangement of the vertical movement component 301, the camera mounting arm 302, the camera mounting plate 303, and the detection camera 305 and the annular light strips 306 thereon can move in the vertical direction. During the food packaging detection process, different types of packaging bags may require different detection heights. Through the vertical movement component 301, the height of the detection camera 305 can be flexibly adjusted, so that the camera can be in the best shooting position, ensuring that the image information of the packaging bag can be clearly collected, and improving the accuracy and applicability of the detection. Specifically, the vertical movement component 301 can adopt a conventional lead screw movement component, and the specific structure will not be elaborated.

[0054] In a specific embodiment, several annular light belts 306 can adopt different types of lights. Cooperating with the detection camera 305, images of the packaging bag under different lights can be obtained, which is convenient for subsequent image processing when needed.

[0055] As Figures 2-3 shown, a light-shielding cover 307 is provided on the lower surface of the camera mounting plate 303. The light-shielding cover 307 and the camera mounting plate 303 cooperate to form a sealed space with an open lower end. A flange is provided at the edge of the upper surface of the placing tray 2033. The shape of the flange is adapted to the light-shielding cover 307 and is spaced from the light-shielding cover 307.

[0056] The light-shielding cover 307 provided on the lower surface of the camera mounting plate 303 and the camera mounting plate 303 cooperate to form a sealed space with an open lower end. During the visual inspection process, external light may interfere with the images collected by the detection camera 305, resulting in problems such as reflection and halo in the images, affecting the accuracy of the inspection. The light-shielding cover 307 can effectively block the entry of external light, creating a relatively stable and uniform lighting environment for the detection camera, making the collected images of the packaging bag clearer and more accurate, and improving the effect of visual inspection.

[0057] As Figure 6 shown, the structure of the loading manipulator 1 is the same as that of the unloading manipulator 4. The loading manipulator 1 includes: a mounting base 101, a first turntable 102 is fixedly connected to the mounting base 101. The rotating part of the first turntable 102 is fixedly connected to one end of a first extension arm 103. The other end of the first extension arm 103 is fixedly connected to a second turntable 104. The rotating part of the second turntable 104 is fixedly connected to one end of a second extension arm 105. The other end of the second extension arm 105 is fixedly connected to a third turntable 106. The working end of the third turntable 106 is fixedly connected to a vertically arranged lifting rod, and the lower end of the lifting rod is fixedly connected to a suction plate 107.

[0058] The loading manipulator 1 is equipped with a first turntable 102, a second turntable 104 and a third turntable 106. The rotating parts of these turntables can drive the corresponding extension arms to rotate. Through the linkage cooperation of the three turntables, the manipulator can achieve flexible movement in multiple angles and directions in three-dimensional space. This design with multiple turntables enables the manipulator to easily reach various positions within the working area and adapt to different loading and unloading position requirements.

[0059] In a specific embodiment, the shape of the suction plate 107 is adapted to the inner shape of the flange of the placing tray 2033. The adapted shape can enable the suction plate 107 to adsorb the packaging bag more stably and minimize the wrinkles generated on the surface of the packaging bag during the transfer process.

[0060] The present application also provides a visual inspection method for food packaging in food processing. The inspection method uses the inspection device as described above and includes the following steps:

[0061] The loading manipulator 1 transfers the packaging bag to the placement plate 2034 of the turntable feeding mechanism 2;

[0062] The support turntable 201 of the turntable feeding mechanism 2 rotates to move the placement plate 2034 with the packaging bag placed thereon to the inspection area of the visual inspection mechanism 3. During the movement, the vibration motor 2032 works to flatten the packaging bag on the placement plate 2034;

[0063] When the placement plate 2034 moves to the inspection area of the visual inspection mechanism 3 and the placement plate 2034 completely falls within the vertical projection of the light-shielding cover 307, the inspection camera 305 acquires the image information of the packaging bag, analyzes and processes the image information, and classifies the packaging bag into a good-quality packaging bag, a defective packaging bag, and a suspicious packaging bag;

[0064] After the inspection is completed, the support turntable 201 rotates to move the placement plate 2034 out of the inspection area of the visual inspection mechanism 3. The unloading manipulator 4 transfers the packaging bag to the good-quality conveyor belt 5, the defective conveyor belt 6, or the buffer box 7 according to the inspection result.

[0065] In a specific embodiment, the analysis and processing of the image information includes:

[0066] Preprocessing the acquired image information, and the preprocessing includes filtering, noise reduction, and enhancement operations;

[0067] Using an edge detection algorithm to extract the edge information of the packaging bag;

[0068] Calculating the number of pixels within the area enclosed by the edge;

[0069] Converting the pixel area to the actual area according to the pre-calibrated conversion relationship between pixels and actual dimensions, so as to obtain the image area of the packaging bag;

[0070] Calculating the ratio of the image area of the packaging bag to the calibrated area of the packaging bag, and taking it as the wrinkle coefficient. The calibrated area of the packaging bag is the actual area when the packaging bag is completely flattened;

[0071] When the wrinkle coefficient exceeds the preset range, the packaging bag is calibrated as a suspicious packaging bag, and no surface feature analysis is performed; the suspicious packaging bag is subsequently re-flattened by a flattening mechanism or manually flattened by a human and then re-inspected;

[0072] When the wrinkle coefficient is within the preset range, surface feature analysis is performed on the packaging bag image, and the surface feature analysis includes surface defect detection and printing quality detection;

[0073] The packaging bag is calibrated as a good product packaging bag or a defective product packaging bag according to the surface feature analysis result. The existing image feature analysis methods that can be used for surface feature analysis will not be elaborated here.

[0074] The steps of performing filtering, noise reduction, and enhancement operations on the image specifically include:

[0075] Assume the original image is I, and the image after preprocessing is I'; filtering and noise reduction can be achieved through convolution operations. Assume the convolution kernel is K, then the filtered image I filtered is expressed as:

[0076] I filtered = I * K, where * represents the convolution operation.

[0077] Image enhancement can be achieved by adjusting the contrast and brightness. The enhanced image I' can be expressed as:

[0078] I' = α·I filtered + β; where α is the contrast enhancement coefficient and β is the brightness enhancement coefficient.

[0079] Specifically, the contrast enhancement coefficient α and the brightness enhancement coefficient β can be selected and configured according to packaging bags of different colors:

[0080] For light-colored packaging bags (such as white, light yellow): Light-colored packaging bags are prone to being confused with the background. The value of α can be appropriately increased (such as α = 1.3) to enhance the contrast and separate the packaging bag from the background; while the light-colored packaging bag itself has a high brightness, the value of β can be appropriately decreased (such as β = 5) to avoid overbrightness.

[0081] For dark-colored packaging bags (such as black, dark blue): The details of dark-colored packaging bags are prone to being ignored. The value of α can be appropriately increased (such as α = 1.5) to enhance the contrast and highlight the details; the value of β for dark-colored packaging bags needs to be appropriately increased (such as β = 20) to enhance the brightness and make the details clearer.

[0082] Compared with colored packaging bags: Colored packaging bags need to maintain the naturalness of the color. An appropriate value of α (such as α = 1.2) can be adopted to avoid color distortion; the value of β for colored packaging bags needs to be adjusted according to the specific color and can usually be maintained at a medium level (such as β = 10).

[0083] The specific steps of using the edge detection algorithm to extract the edge information of the packaging bag include: Using the edge detection algorithm (such as Sobel, Canny, etc.) to extract the edge information of the packaging bag. The image E after edge detection is expressed as:

[0084] E = EdgeDetect(I');

[0085] Among them, EdgeDetect is the edge detection function, and the edge intensity is represented by the gradient magnitude G. G x and G y are the gradients of the image in the x and y directions respectively, which can be calculated through convolution kernels (such as Sobel operator, Prewitt operator or Roberts operator). The gradient magnitude G is an important basis for judging whether a pixel belongs to the edge. Usually, a threshold T is set. When G>T, this pixel is considered to be part of the edge.

[0086] Assume that the binary image after edge detection is B, where the pixel value of the edge area is 1 and the pixel value of the non-edge area is 0. Then the number of pixels P in the edge area is expressed as:

[0087] where W and H are the width and height of the image respectively.

[0088] The above-mentioned conversion of the pixel area into the actual area according to the pre-calibrated conversion relationship between pixels and actual dimensions, so as to obtain the image area of the packaging bag, specifically includes:

[0089] According to the pre-calibrated conversion relationship between pixels and actual dimensions, the pixel area is converted into the actual area. Assume that the conversion coefficient between pixels and actual dimensions is D (unit: pixel / mm²), then the image area A of the packaging bag p is expressed as:

[0090] A p =P·D.

[0091] The calibrated area of the packaging bag (the actual area when it is fully flattened) is A r , then the wrinkle coefficient C can be expressed by the following formula:

[0092]

[0093] In a specific embodiment, the wrinkle coefficient where A p is the image area of the packaging bag, A r is the calibrated area of the packaging bag,

[0094] When C min <C<C max at this time, surface feature analysis is performed on the packaging bag image. On the contrary, when the wrinkle coefficient is not within the preset range, it is marked as a suspicious packaging bag;

[0095] C min =a0 + a1M + a2MA r , where a0, a1, and a2 are empirical coefficients, and M is the packaging bag material coefficient;

[0096] Cmax = b0 + b1M + b2MA r , where b0, b1, and b2 are empirical coefficients;

[0097] The ease of wrinkling of packaging bags made of different materials varies under the same conditions. The material coefficient can quantify this difference. Linking the preset range with the material coefficient can more accurately reflect the wrinkling characteristics of packaging bags made of different materials and avoid judgment errors caused by ignoring the material factor;

[0098] The actual area size of the packaging bag will affect the formation and manifestation of wrinkles. Generally speaking, packaging bags with a larger area are more likely to be affected by external forces during production, transportation, or storage, thus increasing the probability of wrinkles appearing. By considering the influence of the actual area on the preset range, the wrinkling situation of packaging bags of different sizes can be judged more accurately.

[0099] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the inventive concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. A visual inspection device for food packaging in food processing, characterized in that, Including: a loading manipulator (1), a rotary table feeding mechanism (2), a vision inspection mechanism (3), an unloading manipulator (4), a good product conveyor belt (5), a defective product conveyor belt (6), and a buffer box (7); The loading manipulator (1) is used to transfer the packaging bag to be tested to the rotary table feeding mechanism (2), and the rotary table feeding mechanism (2) is used to transfer the packaging bag to be tested to the detection area of the vision inspection mechanism (3); the vision inspection mechanism (3) is used to perform vision inspection on the packaging bag and distinguish the packaging bag into a good product packaging bag and a defective product packaging bag according to the inspection result; specifically, the vision inspection mechanism (3) is used to collect the image information of the packaging bag, and then preprocess the collected image information, and the preprocessing includes filtering, noise reduction, and enhancement operations; and use an edge detection algorithm to extract the edge information of the packaging bag; calculate the number of pixels in the area enclosed by the edge; then convert the pixel area into the actual area according to the pre-calibrated conversion relationship between pixels and actual dimensions, so as to obtain the image area of the packaging bag; and calculate the ratio of the packaging bag image area to the calibrated area of the packaging bag, and use it as the wrinkle coefficient, and the calibrated area of the packaging bag is the actual area when the packaging bag is completely flattened; finally, judge whether the wrinkle coefficient exceeds the preset range, if so, judge it as a suspicious packaging bag, otherwise perform surface feature analysis on the packaging bag image, and calibrate the packaging bag as a good product packaging bag or a defective product packaging bag according to the surface feature analysis result; The unloading manipulator (4) transfers the packaging bag to the good product conveyor belt (5) or the defective product conveyor belt (6) or the buffer box (7) according to the inspection result of the vision inspection mechanism (3).

2. The visual inspection device for food packaging in food processing according to claim 1, wherein The rotary table feeding mechanism (2) includes: a support turntable (201), an installation disk (202) is horizontally arranged coaxially on the support turntable (201), and a plurality of groups of placement components (203) are arranged on the upper surface of the installation disk (202), and a vibration motor (2032) is arranged in the middle of the placement component (203); The placement component (203) includes: a mounting plate (2031), the mounting plate (2031) is fixedly connected to the mounting disk (202) through a fixing bolt (2037), a vibration motor (2032) is arranged in the middle of the mounting plate (2031), a placement disk (2033) is abutted against the upper part of the vibration motor (2032), four corners of the lower surface of the placement disk (2033) are vertically provided with connecting rods (2034), mounting holes matched with the connecting rods (2034) are opened on the upper surface of the mounting plate (2031), pin holes (2035) communicated with the mounting holes are opened on the side wall of the mounting plate (2031), and pins for fixing the connecting rods (2034) are threadedly connected in the pin holes (2035).

3. The visual inspection device for food packaging in food processing according to claim 2, wherein, A mounting groove is formed in the middle of the upper surface of the mounting plate (2031). A first shock-absorbing layer (2036) is provided on the circumferential side wall of the mounting groove. The vibration motor (2032) is in interference fit with the first shock-absorbing layer (2036). A second shock-absorbing layer (2038) is directly provided between the upper surface of the vibration motor (2032) and the placement plate (2033).

4. The visual inspection device for food packaging in food processing according to claim 2, characterized in that, The visual inspection mechanism (3) includes: a vertical movement assembly (301). A camera mounting arm (302) is provided on the vertical movement assembly (301). A camera mounting plate (303) is provided at the end of the camera mounting arm (302). A light strip plate (304) is provided on the lower surface of the camera mounting plate (303). A detection camera (305) and a plurality of annular light strips (306) are provided on the lower surface of the light strip plate (304). The detection camera (305) is located in the middle of the light strip plate (304). The plurality of annular light strips (306) are concentrically distributed with the axis of the detection camera (305) as the center of the circle.

5. A food packaging visual inspection device in food processing according to claim 4, characterized in that, A light-shielding cover (307) is provided on the lower surface of the camera mounting plate (303). The light-shielding cover (307) and the camera mounting plate (303) cooperate to form a closed space with an open lower end. A flange is provided at the edge of the upper surface of the placement plate (2033). The shape of the flange is adapted to the light-shielding cover (307) and is spaced from the light-shielding cover (307).

6. The visual inspection device for food packaging in food processing according to claim 5, characterized in that, The structure of the loading manipulator (1) is the same as that of the unloading manipulator (4). The loading manipulator (1) includes: a mounting base (101). A first turntable (102) is fixedly connected to the mounting base (101). The rotating part of the first turntable (102) is fixedly connected to one end of a first extension arm (103). The other end of the first extension arm (103) is fixedly connected to a second turntable (104). The rotating part of the second turntable (104) is fixedly connected to one end of a second extension arm (105). The other end of the second extension arm (105) is fixedly connected to a third turntable (106). The working end of the third turntable (106) is fixedly connected to a vertically arranged lifting rod. The lower end of the lifting rod is fixedly connected to an adsorption plate (107).

7. The visual inspection device for food packaging in food processing according to claim 6, characterized in that, The shape of the adsorption plate (107) is adapted to the shape of the inner side of the flange of the placement plate (2033).

8. A visual inspection method for food packaging in food processing, characterized in that, The detection method uses the detection device according to any one of claims 5-7 and includes the following steps: The loading manipulator (1) transfers the packaging bag to the placement plate (2034) of the turntable feeding mechanism (2). The support turntable (201) of the turntable feeding mechanism (2) rotates to move the placement plate (2034) with the packaging bag placed thereon to the detection area of the visual inspection mechanism (3). During the movement, the vibration motor (2032) works to flatten the packaging bag on the placement plate (2034). The placement tray (2034) is moved to the detection area of the vision detection mechanism (3). After the placement tray (2034) completely falls within the vertical projection of the light-shielding cover (307), the detection camera (305) acquires the image information of the packaging bag, analyzes and processes the image information, and classifies the packaging bags into qualified packaging bags, defective packaging bags, and suspicious packaging bags; After the detection is completed, the support turntable (201) rotates to move the placement tray (2034) out of the detection area of the vision detection mechanism (3). The blanking manipulator (4) transports the packaging bag to the qualified product conveyor belt (5) or the defective product conveyor belt (6) or the buffer box (7) according to the detection result.

9. A method for visual inspection of food packaging in food processing according to claim 8, characterized in that, The analysis and processing of the image information includes: Preprocessing the acquired image information, and the preprocessing includes filtering, noise reduction, and enhancement operations; Using an edge detection algorithm to extract the edge information of the packaging bag; Calculating the number of pixels within the area enclosed by the edges; Converting the pixel area into the actual area according to the pre-calibrated conversion relationship between pixels and actual dimensions, so as to obtain the image area of the packaging bag; Calculating the ratio of the packaging bag image area to the calibrated area of the packaging bag, and taking it as the wrinkle coefficient. The calibrated area of the packaging bag is the actual area when the packaging bag is completely flattened; When the wrinkle coefficient exceeds the preset range, the packaging bag is designated as a suspicious packaging bag, and no surface feature analysis is performed; When the wrinkle coefficient is within the preset range, surface feature analysis is performed on the packaging bag image, and the surface feature analysis includes surface defect detection and printing quality detection; Designating the packaging bag as a qualified packaging bag or a defective packaging bag according to the surface feature analysis result.

Images