The work aims to construct an image detection system to detect the punching size and edge length of 0402 component packaging paper tape, which is regarded as the basis for quality inspection of stamping products and for observation, judgment and adjustment of mold. Firstly, CCD was used to pick up paper tape image and cooperate with motor to drive the platform to move. Secondly, the longest edge size was measured by means of binarization algorithm and morphological interpolation algorithm. Thirdly, LabVIEW programming was used to control image picking and processing. Fourthly, the dimension or edge, average length value, 3σ, maximum/minimum value and difference value, comprehensive process capability index and so on were displayed through man-machine interface to confirm the product quality. The experimental results indicated that, 10 pieces of paper tape could be measured by the image detection system each time, while the time for each piece was 64 s. The maximum positioning error was 0.005 mm, and the maximum error percentage was 3.93%. When compared with the national standards, the position and length of the edge of paper tape found in the edge detection could be used as the basis for adjusting or replacing the mold. The image detection system can replace manual measurement, reduce the error of artificial measurement, save time and improve production efficiency.