Locating the tiny insulator defect object with complex backgrounds in high-resolution aerial images is a challenging task. In this paper, we propose a novel method which cascades detection and segmentation networks to identify the defect from the global and local two levels: (1) The improved Faster R-CNN is carried out to capture both defects and insulators in the entire image. ResNeXt-101 is adopted as the feature extraction network so as to fully extract features, and Feature Pyramid Network (FPN) is built to enhance the ability of detecting small targets. In addition, the Online Hard Example Mining (OHEM) training strategy is applied to solve the imbalance problem of positive and negative samples. (2) All the detected insulators are extracted and fed into the improved U-Net network to futher inspect at pixel level, we utilize the pre-trained ResNeXt-50 as the encoder of U-Net, incorporate an attention module, Spatial and Channel Squeeze & Excitation Block (SCSE), into the decoding path to highlight the meaningful information. A hybrid loss which merges binary cross entropy (BCE) loss and dice coefficient loss is designed to train our network for figuring out the class imbalance issue. The missed detection can be greatly reduced with the combination of two modified network, which makes comprehensive use of the original map information and local information. On the test set of actual images, the insulator defect recognition precision and recall of the cascade network is 91.9% and 95.7%, exhibiting strong robustness and accuracy.

No abstract

No abstract

In this paper, asymmetric transmission with extremely high contrast for linear polarization is proposed in chiral split ring resonators. Results show that only specific cross-polarization can pass through the structure with a high power transmission efficiency of 0.9 (amplitude transmission 0.95), which is much higher than those reported before in infrared region. The contrast between the two cross-polarizations is also very large (larger than 24.5 dB). Another merit of our scheme is that the operation wavelength is broadband, and can shift linearly to achieve asymmetric transmission over a range of 330 nm by the tuning of a single structural parameter. These excellent features of asymmetric transmission make our scheme very promising in practical applications, such as integrated optical diodes.

This paper describes a fault diagnosis expert system for cement kiln developed in the way of integrating the new theories and methods of artificial intelligence and network technology with related production technology . The system can give online fault diagnosis and has features of simple network interface, excellent openness and easy expansibility. The design of the system layout, database, knowledge base and reasoning engine is presented in detail. The experiment application in a factory shows its high adaptability and the wide prospect of the system.