Research And Application Of Coke Oven Number Recognition Based On Deep Learning

The identification of the coke oven number has important practical significance of the production of the coke industry.Coke oven number recognition usually uses encoding and decoding methods,laser ranging sensor methods,radio frequency methods,etc,which have disadvantages such as high cost and poor generalization performance.In recent years,with the development of deep learning technology in the field of computer vision,applications such as target classification,detection,and recognition have made significant progress.Based on this,this article introduces deep learning technology to identify the furnace number of the coke oven,the main content includes furnace number detection and furnace number recognition.In the furnace number detection,the traditional methods such as morphological operation,edge detection,and contour detection are abandoned,and the CTPN algorithm is introduced to detect the furnace number.At the same time,the CTPN anchor design is improved according to the scene.In the image of the number image,the scale of the furnace number area is small,and the matching preset anchor frame is selected.Experiments show that,without reducing the detection accuracy,the detection time is reduced to 1/3 of the original,and the detection accuracy is 98%.in the heat number recognition,the traditional projection transformation and template matching methods are abandoned,and the CRNN algorithm is introduced.To identify the furnace number,the experimental results show that the recognition accuracy rate is 97%,and the average time to identify a single furnace number is 12 ms,and the detection efficiency is significantly higher than that of the projection transformation method.On the basis of the above-mentioned oven number detection and identification algorithm,combined with the actual production needs of the coke plant,a set of coke oven oven number identification software was designed and developed and used in the coke cake temperature management system.Through the use of the system,it shows the effectiveness of the heat identification algorithm designed in this paper.