Research On The Application Of Object Detection Based On Deep Learning In Intelligent Manufacturing

In recent years,deep learning has been developed,and many fruitful results have been obtained in machine vision,text writing,speech processing and medical image detection.among the different types of deep learning neural networks,among which convolutional neural networks have been most deeply studied because of their very powerful ability to extract image features,which is not only reflected in the prosperity of theoretical research,but also has a great influence in commercial activities and industrial production.as the basis of image analysis,object detection is an important research direction in the field of machine vision.object detection is to accurately locate the location area where the detection object is located in the image and judge the category to which the detection object belongs.it is widely used in machine inspection,product detection,autopilot and human-computer interaction.In the process of intelligent manufacturing,the identification,positioning and detection of object products on the production line is the basic requirement of precision machining.The application of deep learning object detection technology in intelligent manufacturing can make the defect detection of intelligent manufacturing vision system faster,character recognition more accurate and object positioning more efficient,and then can achieve high precision,visual guidance and accurate control and other technology production activities.In this thesis,the object detection algorithm of deep learning is studied from the point of view of intelligent manufacturing.to solve these problems,a object detection algorithm based on separable convolution is proposed.introducing deep learning object detection technology into the production and manufacturing monitoring link of intelligent manufacturing input,judging the quality of the product by detecting the image of the product is of great significance to the functionality and flexibility of the intelligent manufacturing system,which is also an important direction of intelligent manufacturing in the future.At present,there is a certain gap between intelligent manufacturing in China and developed countries,and the goalof introducing deep learning to detect intelligent manufacturing is the key to change this gap in intelligent manufacturing in China.In the process of upgrading industrial manufacturing to intelligent manufacturing,intelligent perception is the key position,and the data collected by the instrument is the source of intelligent manufacturing perception,so it is very important to make the instrument intelligent.Based on separable convolution,this thesis uses the Mobile Net-YOLO v3 of object detection algorithm proposed by patrol robot collocation to realize the intelligent data acquisition of instrument.Experiments show that the proposed method can accurately and quickly collect instrument data,and the accuracy of indicator recognition is over 99%.