With the improvement of people's health awareness and quality of life, the demand for sanitary products is growing. However, as an important industrial material, there are technical and environmental restrictions in the current production of non-woven fabrics, resulting in defects that cannot be detected in time and affect product quality. In order to solve this problem, the team has developed two core technologies, the defect detection algorithm based on feature recognition and the defect classification algorithm based on artificial intelligence, which is the first in the industry, and has launched three products, namely, high-speed defect detection equipment based on feature recognition, high-speed micro defect detection equipment based on artificial intelligence, artificial intelligence defect detection and recognition integrated machine and its supporting cloud platform, with industry-leading processing speed and low computing consumption, adapt to high-speed production lines, support defect classification and identification, and control the cost within 50,000 yuan. After optimization and debugging, the detection rate can reach 98%, and the detection speed is far faster than similar products.