Bridge crack detection is essential to ensure bridge safety. The introduction of deep learning technology has made it possible to detect bridge cracks automatically and accurately. In this study, the Inception-Resnet-v2 algorithm was systematically improved and applied to the real-time detection of bridge cracks. We propose an end-to-end bridge crack detection model based on a convolutional neural network. This model combines the advantages of Inception convolution and residual networks, broadening the network width and alleviating the training problem of the deep network. The calculation speed is improved while still ensuring accuracy. Multi-scale feature fusion enables the network to extract contextual information of different scales, which improves the accuracy of crack recognition. The GKA (K-means clustering method based on a genetic algorithm) realizes the accurate segmentation of the target area, greatly enhances the clustering effect, and effectively improves the detection speed. In this model, large fracture datasets are used for training and testing without pre-training. The experimental results show that the performance of this method was improved in all aspects: accuracy, 99.24%; recall, 99.03%; F-measure, 98.79%; and FPS(Frames Per Second), 196.

Multispectral pedestrian detection based on deep learning can provide a robust and accurate detection under different illumination conditions, which has important research significance in safety. In order to reduce the log-average miss rate of the object under different illumination conditions, a new one-stage detector suitable for multispectral pedestrian detection is proposed. First, in order to realize the complementarity between the information flows of the two modalities in the feature extraction stage to reduce the object loss, a low-cost cross-modality feature complementary module (CFCM) is proposed. Second, in order to suppress the background noise in different environments and enhance the semantic information and location information of the object, so as to reduce the error detection of the object, an attention-based feature enhancement fusion module (AFEFM) is proposed. Thirdly, through the feature complementarity of color-thermal image pair and the multi-scale fusion of depth feature layer, the horizontal and vertical multi-dimensional data mining of parallel deep neural network is realized, which provides effective data support for object detection algorithm. Finally, through the reasonable arrangement of proposed modules, a robust multispectral detection framework is proposed. The experimental results on the Korea Advanced Institute of Science and Technology (KAIST) pedestrian benchmark show that the proposed method has the lowest log-average miss rate compared with other state-of-the-art multispectral pedestrian detectors, and has a good balance in speed and accuracy.

With the exploration and development of marine resources, deep learning is more and more widely used in underwater image processing. However, the quality of the original underwater images is so low that traditional semantic segmentation methods obtain poor segmentation results, such as blurred target edges, insufficient segmentation accuracy, and poor regional boundary segmentation effects. To solve these problems, this paper proposes a semantic segmentation method for underwater images. Firstly, the image enhancement based on multi-spatial transformation is performed to improve the quality of the original images, which is not common in other advanced semantic segmentation methods. Then, the densely connected hybrid atrous convolution effectively expands the receptive field and slows down the speed of resolution reduction. Next, the cascaded atrous convolutional spatial pyramid pooling module integrates boundary features of different scales to enrich target details. Finally, the context information aggregation decoder fuses the features of the shallow network and the deep network to extract rich contextual information, which greatly reduces information loss. The proposed method was evaluated on RUIE, HabCam UID, and UIEBD. Compared with the state-of-the-art semantic segmentation algorithms, the proposed method has advantages in segmentation integrity, location accuracy, boundary clarity, and detail in subjective perception. On the objective data, the proposed method achieves the highest MIOU of 68.3 and OA of 79.4, and it has a low resource consumption. Besides, the ablation experiment also verifies the effectiveness of our method.

Bridge crack is one of the critical optical and visual information to judge the health state of bridges. The bridge crack detection methods based on artificial intelligence are essential in this field, but the current approaches are not satisfactory in terms of speed and accuracy. This study proposes a novel multi-scale crack detection network, called MSCNet, comprising a texture enhancement mechanism and feature aggregation to enhance the visual saliency of the objects in the background for bridge crack detection. We use Res2Net as the backbone network to improve the depth information expression ability of the cracks itself. Because the edge property of bridge cracks is prominent, to make full use of this visual feature, we use a texture enhancement module based on group attention to capturing the detailed information of cracks in low-level features. To further mine the depth information of the network, we use a cascade fusion module to capture crack location information in high-level features. Finally, to fully utilize the characteristic information of the deep network, we fuse the low- and high-level features to obtain the final crack prediction. We evaluate the proposed method compared with other state-of-the-art methods on a large-scale crack dataset. The experimental results demonstrate the effectiveness and superiority of the proposed method, which achieves a precision of 93.5%, recall of 94.2%, and inference speed of over 63 FPS.

To solve the problem of low detection accuracy of small objects in UAV optical remote sensing images due to low contrast, dense distribution, and weak features, this paper proposes a small object detection method based on feature alignment of candidate regions is proposed for remote sensing images. Firstly, AFA-FPN (Attention-based Feature Alignment FPN) defines the corresponding relationship between feature mappings, solves the misregistration of features between adjacent levels, and improves the recognition ability of small objects by aligning and fusing shallow spatial features and deep semantic features. Secondly, the PHDA (Polarization Hybrid Domain Attention) module captures local areas containing small object features through parallel channel domain attention and spatial domain attention. It assigns a larger weight to these areas to alleviate the interference of background noise. Then, the rotation branch uses RROI to rotate the horizontal frame obtained by RPN, which avoids missing detection of small objects with dense distribution and arbitrary direction. Next, the rotation branch uses RROI to rotate the horizontal box obtained by RPN. It solves the problem of missing detection of small objects with dense distribution and arbitrary direction and prevents feature mismatch between the object and candidate regions. Finally, the loss function is improved to better reflect the difference between the predicted value and the ground truth. Experiments are conducted on a self-made dataset. The experimental results show that the mAP of the proposed method reaches 82.04% and the detection speed reaches 24.3 FPS, which is significantly higher than that of the state-of-the-art methods. Meanwhile, the ablation experiment verifies the rationality of each module.

The remote sensing images in large scenes have a complex background, and the types, sizes, and postures of the targets are different, making object detection in remote sensing images difficult. To solve this problem, an end-to-end multi-size object detection method based on a dual attention mechanism is proposed in this paper. First, the MobileNets backbone network is used to extract multi-layer features of remote sensing images as the input of MFCA, a multi-size feature concentration attention module. MFCA employs an attention mechanism to suppress noise, enhance effective feature reuse, and improve the adaptability of the network to multi-size target features through multi-layer convolution operation. Then, TSDFF (two-stage deep feature fusion module)deeply fuses the feature maps output by MFCA to maximize the correlation between the feature sets and especially improve the feature expression of small targets. Next, the GLCNet (global-local context network) and the SSA (significant simple attention module) distinguish the fused features and screen out useful channel information, which makes the detected features more representative. Finally, the loss function is improved to truly reflect the difference between the candidate frames and the real frames, enhancing the network’s ability to predict complex samples. The performance of our proposed method is compared with other advanced algorithms on NWPU VHR-10, DOTA, RSOD open datasets. Experimental results show that our proposed method achieves the best AP (average precision) and mAP (mean average precision), indicating that the method can accurately detect multi-type, multi-size, and multi-posture targets with high adaptability.