On-line real-time detection system for gray fabric flaw based on machine vision

The utility model relates to an on-line real-time detection system for gray fabric flaw based on machine vision, which comprises a cross beam device. A light source device and one or a plurality of industrial cameras are fixedly arranged on the cross beam device, an imaging area of one industrial camera or an imaging area formed by splicing the plurality of industrial cameras covers a to-be-detected area through which detected gray fabric can pass during spinning, the industrial camera(s) is/are connected with an image acquiring, processing and analyzing device, the image acquiring, processing and analyzing device is connected with a control unit used for controlling a warp knitting machine or a spandex machine, and the control unit is connected with a flaw position area display device. The on-line real-time detection system is simple in structure, low in power consumption and lower in use and maintenance cost, and is safe and reliable, on-line operation of workers is not affected, and the product quality and the efficiency of the warp knitting machine or the spandex machine are greatly improved; therefore, the on-line real-time detection system is applicable for being widely popularized during gray fabric production of the warp knitting machine or the spandex machine and is adapted to the development trend of automatic and unmanned spinning.