Step-by-step interactive contour recognition method based on deep learning model

• Main Authors: PAN YONGCAI, LIN CHUAN, GU JIAHONG, ZHANG ZHENGUANG, QIAO YAKUN, WU HAICHEN, PANG CHEN, CHEN YONGLIANG, WEI YANXIA, XIE ZHIXING, LI FUZHANG
• Format: Patent
• Language: Chinese; English
• Published: 19.11.2021
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention aims to provide a step-by-step interactive contour recognition method based on a deep learning model, and the method comprises the following steps: A, constructing a deep neural network structure which specifically comprises a coding network and a decoding network; wherein the coding network is a VGG16 network and is divided into five stages of S1, S2, S3, S4 and S5 by taking pooling layers as boundaries, the decoding network is divided into four stages, and each stage is provided with a sub-decoding network M and a sub-decoding network N; B, inputting an original image into the VGG16 network, and sequentially carrying out five-stage convolution processing; and C, respectively inputting the convolution results of the first convolution layer and the second convolution layer in the five stages S1, S2, S3, S4 and S5 of the VGG16 network into the sub-decoding networks M and N, decoding through the double-decoding network, and finally carrying out fusion processing to obtain a final output contour. T