Due to the layout diversity of metering equipment in electric marketing on-site operations,there is a significant recognition error.Therefore,a machine vision-based method of identifying rule-violated behaviors in electric marketing on-site operations is designed.Using machine vision technology,the on-site operation images of electric marketing were effec-tively processed and features closely related to rule-violated behaviors were extracted.Then the identification of on-site rule-violated behaviors was constructed to achieve the identification.The experimental results showed that the proposed method facilitated accurate identification of rule violations from the images with low error level(within 10 pixels for most action samples and lower than 25 pixels in total).
关键词
机器视觉/电力营销/计量现场/现场作业/违规行为识别
Key words
machine vision/electric marketing/metering site/on-site operation/identification of rule-violated behavior
维普
万方数据