Digital Method And Edge Information Extraction In CCD Scanning Based Defects Inspection

The cleanliness is one of most important indexes for plastic film product, which must be under control. How to improve the cleanliness of plastic film product is an important problem, which manufacturers of plastic film are faced with. To solve the problem, in the thesis, inspecting system of defect in plastic film and correlative technologies were studied. The existing problems in the technology field were analyzed on the basis of extensive reading for data and investigation. The whole scheme design for inspecting system of defects in plastic film was achieved, in which the photoelectric conversion device linear array CCD technology was adopted.The inspecting system of defects in plastic film based on DSO obtains signals of defects in plastic film with irradiation of parallel light by using linear array CCD control circuit. The 1-D defect signals were collected and transferred by using virtual oscillograph DSO-2902, and the 1-D defect signals were analyzed and processed by using computer. Thereby, the defects were precisely figure out.First defect was inexactly located by using mult-resolution. The algorithm for inspecting of defects in plastic film was put forward, and the signals output from linear array CCD were directly transferred to memory units of computer appointed by using virtual oscillograph DSO-2902. The data of signals memorized in computer were processed by using software system of data collecting. Thereby, the inspection of precise size and location of defects were achieved. In order to reduce measurement error, the subpixel edge location of measurement system needs to be achieved. In the thesis, three edge inspecting methods were put forward: linear fitting, symmetry fitting and curve fitting, and the three methods all achieved subpixel edge location. Contrast experiments and quantity analysis show the curve fitting is quicker, better precision, least error...