Research On Defects Detection And Recognition Based On Machine Vision

With the increasing automation of industrial production,while ensuring production efficiency,defects in products still plague many industrial manufacturers.In addition,as people's living standards have gradually improved,consumers' demands for product quality are also increasing.At present,most manufacturers still use manual methods to detect and identify defective products.This not only lowers production efficiency,but also easily leads to missed inspections.It wastes a lot of manpower and material resources.In order to improve the problems caused by manual quality inspection,the use of intelligent detection methods such as machine vision and pattern recognition to replace manual methods in the production line quality monitoring has become an inevitable trend in the development of industrial production.The application of optical components in daily production and life is increasing,such as various camera lenses used in life and coated lenses for experimental research.In the face of so many applications,manufacturers must strictly monitor and control the defects during the production process.Due to defects in production and processing,or scratches caused by collisions and friction during transportation,various flaws appear on the surface of the lens,which seriously affects product quality.According to the above situation,this paper aims to improve the efficiency and accuracy of the lens detection process.The convolutional neural network method is applied to the detection and recognition of lens surface defects.We design a six-layer convolutional neural network to realize the training and detection of samples,and use the two indicators of accuracy and recall rate to quantitatively analyze and evaluate the algorithm of this paper.Finally,we analyze and study the experimental results.The convolutional neural network method was used to detect the lens with flaws in the recognition surface with high efficiency and accuracy of detection and recognition.It shows the effectiveness of the research method and its significance.