Laser Directional Energy Deposition Sputtering Counting Method of Full Convolution Neural Network

Zhengxiong Li, et al.

Abstract

The invention provides a laser directional energy deposition sputtering counting method of a full convolution neural network. Collecting laser directional energy deposition images, finding all contours in each image by using a topological structure analysis method, establishing a minimum rectangular frame for the contours, segmenting the images in the minimum rectangular frame, labeling labels, and constructing a training set; building a full convolution neural network, inputting the image and the label into the network for prediction, building a loss function and training to a Nash equilibrium state optimization network; collecting images to be detected, carrying out morphological processing, finding the outline of each image by using a topological structure analysis method, establishing a minimum rectangular frame, segmenting the image in the minimum rectangular frame, and inputting network prediction: if sputtering is carried out, the pixel is reserved; otherwise, setting the color as black; and (3) carrying out graying and gradient pixel processing on the image, finding the outline by adopting a watershed algorithm, and counting the number of the outlines to obtain the sputtering number. The invention identifies and counts the sputtering in the laser directional energy deposition process, and is beneficial to the regulation and control of the manufacturing process.

More information can be found here