To meet the requirement of industrial online monitoring on the stability, real-time performance, accuracy and interference resistance of computer vision algorithms, in light of the specific circumstances of the real-time monitoring system for five-degrees-of-freedom of the moving parts of the 200MN extruder, a fast and precise location method of the laser beam center point in complex industrial environment was proposed on the basis of the existing methodologies. First, the Canny algorithm was employed to realize the binary conversion of the edge of preprocessed image, and those, among the resulting edges, unsuitable for spot file were screened out. Then, an improved fast Hough algorithm for conversion was applied to preliminarily identify the approximate location and radius, and the spot effective area was set according to the Hough conversion results. At last, the coordinates of laser beam center point of sub-pixel precision were obtained by fitting the least square ellipse in the effective area. The experimental study indicated that, in a jamming environment, it took less than 50 ms for single time detection with the precision of up to 0.15 pixels. This method can realize the high-precision location of the laser beam center point under strong jamming, and, compared with traditional methods, it is featured by better precision, real-time performance and adaptability.