The invention relates to the technical field of
yarn breakage detection of
textile mills, in particular to an intelligent
yarn breakage detection method and
system based on vision, hardware of the method comprises a
spinning machine, two lasers, n cameras, a PLC, a breaker and the like, and
software comprises seven modules; a
laser device emits
laser to
yarn, an industrial camera collects
bright spot images, video acquisition is compatible with RTSP / HTTP, multi-thread acquisition is achieved, and after image preprocessing, yarn breakage + interference / shielding / camera offset /
laser device
anomaly detection (a threshold value can be dynamically corrected) and yarn breakage confirmation, a
signal is sent to a PLC. Based on the 800 * 448 resolution, the
system has the 150-frame / s real-
time processing capability (limited by an industrial
personal computer), the missing report rate is 0, and the
false alarm rate is lt; the method has the advantages that the method can give an alarm within 2% and 3-5s, the experimental accuracy is 99.106%, the
false alarm rate is 0.09%, the yarn loss is greatly reduced, the expandability is good, and the method is superior to a
photoelectric sensor scheme.