CN-122015951-A - System and method for detecting production links of fortune bags

CN122015951ACN 122015951 ACN122015951 ACN 122015951ACN-122015951-A

Abstract

The invention discloses a system and a method for detecting a production link of a foodbag, which belong to the technical field of food production and comprise a camera shooting unit, an environment and equipment state sensing unit, an alarm unit, a data processing unit and a data processing unit, wherein the camera shooting unit comprises a high-definition color industrial camera, an infrared thermal imaging camera and a global shutter camera, the environment and equipment state sensing unit comprises a temperature sensor, a humidity sensor, a weight sensor and a gas flow rate sensor, the alarm unit comprises a three-color tower lamp, an RGB LED lamp strip, a regional loudspeaker, a directional sound, a personnel wireless earphone and a visual terminal, and the data processing unit comprises an edge computing gateway and a central server. The system comprises a camera unit, an environment and equipment state sensing unit, an alarm unit and a data processing unit, wherein the camera unit can acquire images of a production link of a bag for analysis, the environment and equipment state sensing unit can acquire various data generated by the production link for analysis, the alarm unit can improve production safety, and the data processing unit comprises an edge computing gateway and a central server and is suitable for factory environments.

Inventors

XU RUI
SHEN SHIMING
GUO LIYI
LI JIAN
SUN SHAOHUA
LU JIAN
CHEN ZHE
SUN JIALI
YANG NAIBIN

Assignees

桂林电子科技大学
北海玖嘉久食品有限公司

Dates

Publication Date: 20260512
Application Date: 20260127

Claims (10)

1. The utility model provides a fortune bag production link detecting system which characterized in that includes: an image pickup unit comprising: a high definition color industrial camera arranged at the station of color identification and detail identification, An infrared thermal imaging camera arranged in the inner stuffing preparation area and the quick freezing area, A global shutter camera arranged at the filling station; an environment and device state sensing unit comprising: a temperature sensor for detecting the inner stuffing cutting and mixing pot interlayer, the interior of the cooking cabinet and the interior of the instant freezer, Humidity sensor set in the stuffing preparing room and cooling area, A weight sensor for detecting the discharge weight of filling materials of the filling machine and the final weight of the blessing bags, A gas flow rate sensor disposed at the vent and at a specific operating point; an alarm unit comprising: The three-color tower lamp is arranged at the station and the production line master console, and is used for representing the current operation state of the equipment, namely an RGB LED lamp band, a regional loudspeaker, a directional sound box, a personnel wireless earphone and a visual terminal; a data processing unit comprising: Edge computing gateway and central server.
2. The system for detecting the production links of the fodders according to claim 1, wherein the stations for identifying the colors and the details comprise squeezing molding, inner stuffing preparation, bean skin baking, stuffing filling and tying, and correspondingly, the resolution of the high-definition color industrial camera is not lower than 200 ten thousand pixels, and the frame rate is above 25 fps; The infrared thermal imaging camera is used for acquiring the central temperature of the slurry and the surface temperature distribution of the product, and the data processing unit judges the temperature uniformity based on the data of the infrared thermal imaging camera; the global shutter camera is used for capturing the action of the filling nozzle and the flowing state of stuffing, and the data processing unit judges whether jelly effect occurs or not based on the data of the global shutter camera.
3. The system of claim 2, wherein the high-definition color industrial camera is installed at the outlet of the frying pan to monitor the color and bubble form of the bean skin, above the press forming device to monitor the integrity of the bean skin, in the vertical direction of the contact area between the bean skin and the soldering iron to ensure the clarity of fonts and no breakage, obliquely above the inlet of special stuffing to ensure the throwing of each bag, in the close range above the tying device to capture the winding mode, tightening degree and tail length of the cucurbit vine, above the wobble plate area to monitor the erection state and spacing of the bags, around the cabinet door of the steaming machine to monitor the surface state of the bags before entering the cabinet, in the inlet of the single freezer to monitor the feeding state and whether deformation exists in the products; the infrared thermal imaging camera is arranged above the chopping pan to monitor the feeding process and the apparent state of the slurry, and is arranged at the side of the discharge hole to monitor the visual appearance of the viscosity of the stuffing; The global shutter camera is arranged around the filling mouth at multiple positions so as to monitor the filling depth of the filling material and whether the filling material touches the bottom of the bag; The temperature sensor comprises a PT100 platinum resistor and a measuring range of-40 ℃ to 150 ℃, wherein the precision in a chopping and mixing area is +/-0.1 ℃ and the precision in a steaming and quick-freezing area is +/-0.5 ℃; the infrared thermal imaging camera is a non-contact infrared thermometer and is used for monitoring the surface temperature of the fried bean skin and the surface temperature of the cooled foodbag; The measurement range of the humidity sensor is 0-100% RH, and the precision is +/-2% RH; The measuring range of the weight sensor is 0-50g, and the precision is +/-0.1 g.
4. A method for detecting a production link of a bag, which is applicable to the system of any one of claims 1 to 3, and is characterized by comprising the following steps: Establishing detection targets of a beancurd sheet production link, an inner stuffing preparation link, a blessing bag forming link, a personnel operation link, a tray arranging and steaming link; Collecting data based on the camera unit, the environment and equipment state sensing unit and the alarm unit for analysis by a data processing unit to determine the monitoring condition of the current link; and outputting improved data according to the monitoring condition.
5. The method for detecting the production link of the foodbag according to claim 4, wherein the detection targets of the production link of the beancurd sheet comprise appearance of the fried beancurd sheet, integrity of the pressed beancurd sheet and operation action; The corresponding processing algorithm comprises: color anomaly detection, namely extracting an ROI (region of interest) area, converting the ROI area into an HSV (hue, saturation and value) space, counting the proportion of pixels conforming to a standard color range, and triggering prompt when the proportion of pixels is lower than a threshold value; Performing defect detection, namely performing pixel-level classification on the bean skin image by adopting a semantic segmentation model based on deep learning, and identifying burnt and damaged areas; and (3) monitoring the action, namely monitoring the frequency and the range of turning the bean skin by an operator through an optical flow method or a background subtraction method, and judging whether the bean skin is operated according to a rule.
6. The method for detecting the production link of the fodder bag according to claim 5, wherein the detection targets of the preparation link of the inner stuffing comprise feeding action and sequence, slurry temperature and apparent viscosity of the stuffing; The corresponding processing algorithm comprises: using a model based on time sequence or a predefined action template for matching, analyzing a monitoring video, judging whether a feeding action occurs or not and judging whether the time point of the feeding action accords with a preset sequence or not; The temperature monitoring and early warning is that the data of the temperature sensor is read in real time and is compared with a preset process curve, and the alarm is given when the temperature exceeds the range; Visual viscosity assisted assessment, which is to provide a qualitative reference by analyzing the speed and morphology of the falling filling in the video when the filling is discharged, and comparing the video with a standard video.
7. The method for detecting the production link of the fodder bag according to claim 6, wherein the detection targets of the formation link of the fodder bag comprise baking characters, filling flying fish roes, filling stuffing and bundling bags; The corresponding processing algorithm comprises: Performing OCR and template matching, namely performing image acquisition on the baked bean skin, performing perspective correction, and then using OCR technology or image template matching to calculate the similarity with standard fonts; Target detection and action analysis, namely identifying flying fish roe and operation hand by using a YOLO or SSD target detection model, and judging whether the input action is completed or not by combining track tracking; The weight is read and judged in real time, wherein a weight sensor is integrated in the filling machine, and data is directly accessed into the data processing unit to carry out real-time judgment; Image segmentation and measurement, namely performing semantic segmentation on the stuffed bean skin bag side view image, identifying a stuffing area and a bean skin bag area, and calculating the bottom position of the stuffing; And (3) key point detection and measurement, namely detecting the winding points, the binding points and tail end points of the cucurbit vine, and calculating the winding number, the distance between the binding points and the bag mouth and the tail length of the detected close-up image after binding.
8. The method for detecting the production link of the foodbag according to claim 7, wherein the detection targets of the swinging plate and the cooking link comprise a standing state, a distance, a state of the front surface of the foodbag entering a cabinet and a cooking temperature/time; The corresponding processing algorithm comprises: the attitude estimation comprises the steps of obtaining a top view of a wobble plate area, and judging the standing state of the blessing bag by using the length-width ratio and the direction of a target detection frame or rotating target detection; Calculating the minimum distance between each bag and the adjacent bag boundary frame after each bag is detected; defect detection, namely judging whether filling leakage or serious deformation exists in the bag image by using a classification CNN model; And (4) checking the compliance of the process parameters, namely reading PLC data of the cooking cabinet and carrying out logic judgment.
9. The method for detecting the production links of the blessing bags according to claim 8, wherein the detection targets of the personnel operation links comprise an operator moving state, an on-duty state and an operation rhythm; The corresponding processing algorithm comprises: personnel detection and tracking, namely continuously tracking the position and track of each worker in a workshop by using Re-ID technology and a multi-target tracking algorithm; classifying the behavior states, namely classifying the gestures of the personnel based on skeleton key point detection or a space-time diagram convolution network; and (3) area intrusion detection, namely defining a dangerous area, and immediately alarming when personnel are detected to enter.
10. The method for detecting a production link of a bag according to claim 9, further comprising: And setting an adaptive learning mechanism, setting hierarchical early warning and interaction logic and setting a feedback mechanism.

Description

System and method for detecting production links of fortune bags Technical Field The invention relates to the technical field of food production, in particular to a system and a method for detecting a production link of a foodbag. Background With the rapid development of the modern food industry, the industrial production of prepared foods and traditional flavor foods has become an important way to meet market demands. Fu bag products, such as fish roe Fu bags, fresh shrimp Fu bags and the like, are increasingly favored by consumers due to their unique appearance, rich mouthfeel and lucky moral meaning. The products are usually prepared by wrapping stuffing containing various food materials by using outer skins (such as bean skins) through fine manual or semi-automatic procedures, and finally finishing processing through links such as stewing, cooling, quick freezing and the like. However, the production process of the foodbag product involves a plurality of fine operation links, and extremely high requirements are put on the operation standardization of operators, the stability of the production environment and the running state of equipment. At present, the bag production line still highly depends on traditional manual monitoring and experience judgment, and lacks systematic and intelligent behavior and state monitoring means, which causes significant problems in production efficiency, product quality stability and production safety. Disclosure of Invention The invention provides a system and a method for detecting a production link of a bag, which are used for solving the problems of production efficiency, product quality stability and production safety of the existing production mode. In order to achieve the above purpose, the technical scheme adopted by the invention is as follows: The system comprises a camera unit, an environment and equipment state sensing unit, a data processing unit and a data processing unit, wherein the camera unit comprises a high-definition color industrial camera arranged at a color identification and detail identification station, an infrared thermal imaging camera arranged at an inner filling preparation area and a quick freezing area, a global shutter camera arranged at a filling station, the environment and equipment state sensing unit comprises temperature sensors for detecting an inner filling chopping pot interlayer, the interior of a steaming cabinet and the interior of an instant freezer, humidity sensors arranged at an inner filling preparation room and a cooling area, weight sensors for detecting the discharge weight of filling materials of a filling machine and the final weight check of the bag, a gas flow rate sensor arranged at a ventilation opening and a specific working point, the alarm unit comprises a three-color tower lamp arranged at the station and the production line master console, an RGB LED lamp band for representing the current operation state of equipment, an area loudspeaker, an oriented sound box, a personnel wireless earphone and a visual terminal, and the data processing unit comprises an edge computing gateway and a central server. Further, the stations for color identification and detail identification comprise squeezing molding, inner stuffing preparation, bean skin baking, stuffing filling and tying, correspondingly, the resolution ratio of the high-definition color industrial camera is not lower than 200 ten thousand pixels, the frame rate is above 25fps, the infrared thermal imaging camera is used for acquiring the central temperature of slurry and the surface temperature distribution of products, correspondingly, the data processing unit is used for judging the temperature uniformity based on the data of the infrared thermal imaging camera, the global shutter camera is used for capturing the action of a filling nozzle and the flowing state of stuffing, and the data processing unit is used for judging whether jelly effect occurs based on the data of the global shutter camera. Further, the high-definition color industrial camera is arranged at the outlet of the frying pan to monitor the color and the bubble shape of the bean skin; the bean curd sheet forming device is arranged above the press forming device to monitor the integrity of bean curd sheets; the infrared thermal imaging camera is arranged above the chopping pot to monitor the feeding process and the apparent state of the slurry, is arranged at the side of the discharge port to monitor the visual appearance of the stuffing, is arranged at the periphery of the filling mouth to monitor the filling depth and the contact of the stuffing with the bag bottom, is arranged above the swaying disc area to monitor the upright state and the distance of the fuzhu bags, is arranged around the steaming cabinet door to monitor the surface state of the fuzhu bags before entering the cabinet, is arranged at the inlet of the single freezing machine to monitor the feeding state and the deformation of