Research And Development Of The Online Detection System For Thread Of Bottle Mouth Based On Machine Vision

Glass bottles play a very important role in maintaining the quality and safety of liquorwith the advantage of good appearance and easy using. As the main part of the liquor bottle,the thread of bottle mouth is even more critical. Considered the low efficiency and reliabilityof manual inspection, we researched and developed a vision-based online detection systembased on the technology of machine vision and PLC control.First, after the related technology at home and abroad is researched, according to thecharacteristics of this project, the overall scheme of the online detection system is designed.After the optical imaging characteristic of the thread of bottle mouth is researched, then theback illumination of LED is adopted. According to the accuracy of the inspection and therequirement of the system, the selection of camera and lens are implemented.Then the image processing methods of image pre-processing and image segmentationare discussed in this paper, the treatment effect of different algorithms applied in the screwthread of bottle mouth are analyzed, and the ideal image processing methods for our systemare presented and realized in the paper. The extracting of contour of the screw thread iscompleted with the median filter, the iteration threshold segmentation and the Robertsoperators.The image differential method and image enhancement of transform domain applied inthe defection of screw thread are discussed, and the defection method of screw thread ofbottle mouth based on connective region is raised. The positioning algorithm based on1DHough transform to locate the detected area calibrated in advance is proposed, the divisionof the detected area is completed, the defect detection of thread images, thread contour andno-thread images are implemented.Finally the online detection system for thread of bottle mouth based on machine visionis developed. The system has been used in the production line, the application proves thealgorithm can meet quite well with the requirement of speed and accuracy, and the systemoperates stable and reliable.