Automatic Recognition of Civil Aviation Maintenance Records Based on Deep Learning

The aviation industry is not only a huge economic engine, but also one of the most important sectors that play a key role in safety and security. The analysis of aircraft maintenance records and engine maintenance records is essential for airlines, airports, aviation authorities and manufacturers to ensure that their products are properly maintained. Although this task is important, an automation solution has not been developed. This paper presents a method to automatically recognize civil aviation maintenance records from images. The main task is to identify aircraft maintenance records, engines and airframe components being processed. To solve this problem, we use the depth learning algorithm, and use ground data for pre training. We test by training the model on the image dataset and using the test dataset to compare with the classification results of human experts.