Compressed sensing (CS) is a popular data compression theory for many computer vision tasks, but the high reconstruction complexity for images prevents it from being used in many real-world applications. Existing end-to-end learning methods achieved real time sensing but lack theory guarantee for robust reconstruction results. This paper proposes a neural network called RootsNet, which integrates the CS mechanism into the network to prevent error propagation. So, RootsNet knows what will happen if some modules in the network go wrong. It also implements real-time and successfully reconstructed extremely low measurement rates that are impossible for traditional optimization-theory-based methods. For qualitative validation, RootsNet is implemented in two real-world measurement applications, i.e., a near-field microwave imaging system and a pipeline inspection system, where RootsNet easily saves 60% more measurement time and 95% more data compared with the state-of-the-art optimization-theory-based reconstruction methods. Without losing generality, comprehensive experiments are performed on general datasets, including evaluating the key components in RootsNet, the reconstruction uncertainty, quality, and efficiency. RootsNet has the best uncertainty performance and efficiency, and achieves the best reconstruction quality under super low-measurement rates.

Object positioning is a basic need for visual robotics in automatic assembly lines. An assembly line requires fast transfer to new object positioning tasks with few or no training data for deep learning algorithms, and the captured visual images usually suffer from partial missing and cropping and environmental lighting interference. These features call for efficient and robust arbitrary shape positioning algorithms under data-scarce and shape distortion cases. To this end, this paper proposes the Random Verify Generalised Hough Transform (RV-GHT). The RV-GHT builds a much more concise shape dictionary than traditional GHT methods with just a single training image. The location, orientation, and scaling of multiple target objects are given simultaneously during positioning. Experiments were carried out on a dataset in an automatic assembly line with real shape distortions, and the performance was improved greatly compared to the state-of-the art methods. Although the RV-GHT was initially designed for vision robotics in an automatic assembly line, it works for other object positioning mechatronics systems, which can be modelled as shape distortion on a standard reference object.

In the pigging operation of an oil pipeline, the traditional pigging localization method can only achieve fixed-point detection and cannot track continuously. To ensure the normal operation of the pigging operation and track and locate the pig in a timely manner, a tracking and localization method of an oil pipeline pig based on a noise reduction autoencoder is proposed. First, using the advantages of U-Net (U-shaped Convolutional Neural Network) multi-scale feature extraction, combined with coding and residual-dense blocks, U-Net combined with Residual Dense Block U-shaped Network (RDBU) was proposed for noise reduction, and then real-time localization was calculated out based on the negative pressure wave localization formula of pigs. The experimental results show that, compared with the traditional method, the SNR of the RDBU is increased by 0.9dB and the root mean square error is reduced by 14.92%. The denoising algorithm in this paper can effectively eliminate the noise of the negative pressure wave signal of pigs.

Carbon fiber-reinforced polymer (CFRP) is a widely-used composite material that is vulnerable to impact damage. Light impact damages destroy the inner structure but barely show obvious change on the surface. As a non-contact and high-resolution method to detect subsurface and inner defect, near-field radiofrequency imaging (NRI) suffers from high imaging times. Although some existing works use compressed sensing (CS) for a faster measurement, the corresponding CS reconstruction time remains high. This paper proposes a deep learning-based CS method for fast NRI, this plugin method decreases the measurement time by one order of magnitude without hardware modification and achieves real-time imaging during CS reconstruction. A special 0/1-Bernoulli measurement matrix is designed for sensor scanning firstly, and an interpretable neural network-based CS reconstruction method is proposed. Besides real-time reconstruction, the proposed learning-based reconstruction method can further reduce the required data thus reducing measurement time more than existing CS methods. Under the same imaging quality, experimental results in an NRI system show the proposed method is 20 times faster than traditional raster scan and existing CS reconstruction methods, and the required data is reduced by more than 90% than existing CS reconstruction methods.

To improve pipeline transportation efficiency, long-distance pipelines need to be cleaned regularly. Pipeline inspection gauges (PIGs) are widely used in pigging operations. The traditional PIG localization method generally has a large error and cannot achieve continuous tracking. To solve the problem of PIG real-time tracking and localization, this paper proposes a method of PIG localization for oil pipelines based on a denoising autoencoder(DAE). The attention deconvolution residual block (ADR) is introduced in the decoder part of the DAE, so that this model can learn and use global information to selectively emphasize key features, suppress useless features, and improve the noise removal performance of the model. After denoising, the negative pressure wave(NPW) signal is normalized, the time delay is extracted, and the PIG is located using the NPW localization formula. Experiments show that this method can effectively achieve real-time PIG localization. Compared with the EMD and DAE positioning methods, the relative positioning accuracy is increased by 3.5 times and 8.4 times, respectively.