Sorting gangue from raw coal is an essential concern in coal mining engineering. Prior to separation, the location and shape of the gangue should be extracted from the raw coal image. Several approaches regarding automatic detection of gangue have been proposed to date; however, none of them is satisfying. Therefore, this paper aims to conduct gangue segmentation using a U-shape fully convolutional neural network (U-Net). The proposed network is trained to segment gangue from raw coal images collected under complex environmental conditions. The probability map outputted by the network was used to obtain the location and shape information of gangue. The proposed solution was trained on a dataset consisting of 54 shortwave infrared (SWIR) raw coal images collected from Datong Coalfield. The performance of the network was tested with six never seen images, achieving an average area under the receiver operating characteristics (AUROC) value of 0.96. The resulting intersection over union (IoU) was on average equal to 0.86. The results show the potential of using deep learning methods to perform gangue segmentation under various conditions.

Stymied by lack of energy storage, power systems cannot evolve easily into complex-information networks such as the Internet. However, under the working hypothesis that predicting power behavior is tantamount to energy storage, or stated differently, ldquoour ability to anticipate energy demand can offset our inability to store energyrdquo we begin to develop new energy delivery networks with more transparent relations, higher efficiency and improved reliability. The concept of an energy Internet is discussed and a possible prototype implementation through human-centered automation is presented.

Immediate and accurate detection of foreign metal objects (FMOs) in clothing products is important for guaranteeing human safety. This article proposes an online detection approach based on deep learning, which is suitable for detecting small FMOs from X-ray images of clothing packages. A conveyor belt X-ray scanning system is developed for image collection. The X-ray images are preprocessed by using the morphological erosion operation to improve the accuracy of FMOs detection. These images are then down-sampled to reduce the computation cost. Feature pyramid network (FPN) is adopted for aggregating feature maps with different resolutions, which proved to be effective for small FMOs detection. The stochastic gradient descent (SGD) is used to optimize a multitask loss. The trained model was tested offline on 200 X-ray images, which achieved precision = 0.999, recall = 0.988, F1-score = 0.993, and AP = 0.946. Compared to original Faster region-based convolutional neural networks (R-CNN), the proposed method significantly improved the performance for small FMOs detection in terms of precision and recall rate.

At present, the High Frequency Electric Resistance Welding(HF-ERW) has been applied in many important industrial fields, whose quality problem has been paid much more attention. The HF-ERW quality is related to the surface burr morphology, however, the traditional weld detection equipment can't detect the HF-ERW morphology well, which seriously affects the HF-ERW quality. In order to realize the real-time detection of HF-ERW quality and optimize the process parameters in welding process, this article primarily discusses the process test of the HF-ERW bead from the parameters of 3D solder bead shape derived from linear laser visual sensor. Firstly, a 3D survey model is established to obtain the 3D welding bead shape based on the mapping between world coordinates and image coordinate systems. Then the weld bead dimensions are calculated from the 3D welding bead shape and analysed to get the correlation to the quality of welding bead. The application system of the sensor for the HF-ERW bead is also set up for experiment. The results of experiments prove that the method can accurately measure the shapes of the 3D and 2D geometries in real-time and the parameters can be used to classify the quality status of HF-ERW process.

Motivated by the low efficiency and poor accuracy of the existing techniques for measuring the depth of a pavement structure, we examine the characteristics of laser speckle and speckle measurement methods in relation to the depth of a pavement structure and propose a set of methods for the 3D measurement of such structures based on laser speckle and digital image processing. First, we analyze the relevant characteristics of laser speckle measurement, and then based on the relevant region and temporal sequence correlation, a faster digital speckle correlation method is used to establish a mathematical relationship between the 3D image and the structural depth. The erosion and dilation technique and mean optimization algorithm are proposed to eliminate the noise and burr of primeval 3D reconstruction. After this, a 3D point cloud of information on the depth of the pavement structure is obtained. Finally, through fitting, a good linear relationship was found between the texture depth obtained by the sand paving method and the system. From measurements using a laboratory calibration platform, it was found that the accuracy of measurement of our method can reach the subpixel level, and the running speed is high, meaning that it can be used for highly efficient and accurate detection of the depth of a road structure.

Superficial defect detection on bridge piers is crucial for assessing the health condition of bridges. A novel ring vision acquisition system based on a climbing operation robot is proposed to aim at the limitations of existing Unmanned Aerial Vehicles (UAVs) and wall-climbing robots. In addition, to address the quadratic computational complexity of the Transformer-based model, a Vision Mamba-based defects detection model, Piers Defects-Mamba (PD-Mamba), is proposed, which contributes to more efficient inferencing the high-resolution defects images. Finally, field experiments were performed at Shouchun Bridge in Anhui Province, China. The results demonstrated that the robotic system could automate acquiring high-precision visual images of bridge piers’ superficial defects. The proposed PD-Mamba achieves 77.97% mean Intersection over Union (mIoU), 88.81% mean Precision (mP), 85.42% mean Recall (mR), and 87.05% mean F1-score (mF1) on the multi-category bridge pier defect dataset, exhibiting state-of-the-art (SOTA) performance, and has global modelling capability and linear computational complexity, which achieves accurate and efficient detection of bridge piers’ superficial defects. It is important to assess the health condition of the bridge.