Dual-channel convolutional neural network for power edge image recognition

Abstract In view of the low accuracy and poor processing capacity traditional power equipment image recognition methods, this paper proposes a method based on dual-channel convolutional neural network (DC-CNN) model random forest (RF) classification. aspect feature extraction, DC-CNN extracts characteristics through two independent CNN models. algorithm, by referring to advantages machine learning incorporating RF, an RF classification deep is proposed. Finally, proposed are used classify images various types equipment. The results show that methods can be effectively applied equipment, they greatly improve rate images.