Dilated-FCOS-based teleoperation mechanical arm target detection method

• Main Authors: HUANG JIALE, GUO JIANWEN, MIAO HANQI, LUO WENTAO, SHI FENGSHUO, SU JIAJIE, LI JIANLIN
• Format: Patent
• Language: Chinese; English
• Published: 30.12.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention relates to a teleoperation target detection method, in particular to a dilated-FCOS-based teleoperation mechanical arm target detection method for improving an FCOS algorithm according to the characteristics of large coverage area and high pixel proportion of a teleoperation mechanical arm, and the method is realized through the following steps: S1, detecting a teleoperation mechanical arm target based on FCOS; s2, optimizing an FCOS target detection algorithm; and S3, optimization is realized through experimental design and result analysis: a dilated-FCOS algorithm experiment and a channel pruning dilated-FCOS algorithm experiment are designed, and an experiment result is analyzed, so that an FCOS target detection algorithm is optimized. In order to solve the problem of large target detection of the teleoperation mechanical arm, a target detection method of the dilated-FCOS teleoperation mechanical arm is provided, and meanwhile, channels corresponding to zoom coefficients which have small infl