Paper Packaging Testing System Based On Machine Vision Research And Application

At present, paper cups packing quality are detected by workers in the paper cupproduction line. Due to human errors, the factors of fatigued, and the differencesbetween individuals, therefore a set of automatic detection system must be developed.this system is of great significance to reducing human’s labor intensity, increasingproductivity, and removing the defective cups.The main work of this paper is to design a detection system of instant noodles’spaper cup. detection system automatically remove the defective cups. Instant noodles’spaper cup production process, there will be a few quality problems: posted outside theoff, externally bonded joints appear to varus and valgus, externally bonded joints toowide or too narrow, the cup of the overall deformation, bottom of the cup appearsperforation, and the bottom of the stains. In order to detect these defects, at the first, thispaper design hardware system of the vision inspection system from the hardwarestructure of the machine. That describes the composition of machine vision components,and hardware components for performance comparison and selection.Secondly, this paper completes the overall design of the software system. Carefullystudied the characteristics and design requirements of the test object, the correspondingdetection algorithms designed for different detection function. That is divided into threedetection modules: joint detection, hole detection, stain detection module. This paperdescribes the process of algorithm design ideas, the difficulties encountered andcompleted. Algorithm processing effects and real-time are verified.Finally, the software system is tested in the laboratory and production line from thereal-time and accuracy. System in the operation of the detector to detect tens ofthousands of paper cups, the detection accuracy rate of99.8%.The deficiencies of thesystem and subsequent improvement are proposed base on test results.