| abstract |
The invention relates to a corrugated paper detection algorithm based on neural network. It mainly solves the problems that the existing recognition algorithms are greatly affected by changes in illumination and tile shape, and the algorithm recognition stability is poor. A corrugated paper detection algorithm based on neural network, comprising the following steps: step 1, obtaining a picture of corrugated paper; step 2, building a neural network structure; step 3, collecting and labeling pictures of corrugated paper; step 4, training a neural network; step 5, sliding detection The whole corrugation; step six, fitting the corrugated position; step seven, counting. The corrugated paper detection algorithm based on neural network is less affected by illumination changes and tile shape changes, and the algorithm has high recognition stability. |