Object detection is the most crucial and challenging task of computer vision and has been used in various fields in recent years, such as autonomous driving and industrial inspection. Traditional object detection methods are mainly based on the sliding windows and the handcrafted features, which have problems such as insufficient understanding of image features and low accuracy of detection. With the rapid advancements in deep learning, convolutional neural networks (CNNs) and vision transformers have become fundamental components in object detection models. These components are capable of learning more advanced and deeper image properties, leading to a transformational breakthrough in the performance of object detection. In this review, we comprehensively review the representative object detection models from deep learning periods, tracing their architectural shifts and technological breakthroughs. Furthermore, we discuss key challenges and promising research directions in the object detection. This review aims to provide a comprehensive foundation for practitioners to enhance their understanding of object detection technologies.

With the advancement of computer technology, dynamic detection of damage to tire in-service has become feasible. However, current methods often struggle with accuracy limitations when confronted with specific working conditions and external factors. To address this challenge, we propose an intelligent recognition system for in-service tire damage driven by Strong Combination Augmentation and Contrast Fusion. The system uses a key feature learning enhancement method to address the problem. It uses the Hough transform and the Perceptual Hash algorithm to perform secondary feature comparison, enabling tire region detection even in low-resolution and high-interference scenarios. To effectively eliminate interference caused by wear, stains, and similar factors, we also introduce an efficient damage detection network called CA-EffNet. This network employs a strategic approach that combines various augmentation techniques and parameters with contrast fusion within a supervised learning framework. By integrating these elements, CA-EffNet expands the feature exploration space and effectively captures key damage features. The results show that the system efficiently achieves real-time detection within just 0.7 s at speeds of up to 15 km/h, meeting strict detection requirements. These results highlight the potential of the system to significantly advance the field of tire damage detection and ultimately contribute to safer roads.

With the development of the marine oil industry, leakage accidents are one of the most serious problems threatening maritime and national security. The spilt crude oil can float and sink in the water column, posing a serious long-term threat to the marine environment. High-frequency sonar detection is currently the most efficient method for identifying sunken oil. However, due to the complicated environment of the deep seabed and the interference of the sunken oil signals with homogeneous information, sonar detection data are usually difficult to interpret, resulting in low efficiency and a high failure rate. Previous works have focused on features designed by experts according to the detection environments and the identification of sunken oil targets regardless of the feature extraction step. To automatically identify sunken oil targets without a prior knowledge of the complex seabed conditions during the image acquisition process for sonar detection, a systematic framework is contrived for identifying sunken oil targets that combines image enhancement with a convolutional neural network (CNN) classifier for the final decision on sunken oil targets examined in this work. Case studies are conducted using datasets obtained from a sunken oil release experiment in an outdoor water basin. The experimental results show that (i) the method can effectively distinguish between the sunken oil target, the background, and the interference target; (ii) it achieved an identification accuracy of 83.33%, outperforming feature-based recognition systems, including SVM; and (iii) it provides important information about sunken oil such as the location of the leak, which is useful for decision-making in emergency response to oil spills at sea. This line of research offers a more robust and, above all, more objective option for the difficult task of automatically identifying sunken oils under complex seabed conditions.

Micropiles are a new type of retaining structure widely used in slope engineering due to their small footprint, low vibration and noise emissions, and simple construction process. This study aims to investigate the dynamic response and failure mechanisms of micropiles used in retaining accumulation landslides under seismic loading through shaking table tests and numerical simulation. The failure process, observed phenomena, and bending moments of micropiles in the test were discussed, and the shear force distribution of micropiles was thoroughly analyzed based on numerical simulation. The findings reveal that the natural frequency of the entire landslide system exhibits a gradual decrease and tends to stabilize under sustained earthquake excitation. The bending moment of micropiles follows an “S” shape, with a larger magnitude at the top and a smaller one at the bottom. Additionally, the shear force distribution exhibits a “W-shaped” pattern. Damage to micropiles mainly includes the flexural shear combination failure at the load-bearing section (which occurs within 1.4–3.6 times the pile diameter above the sliding surface) and the shear failure near the sliding surface. This study provides significant insights into the strengthening mechanisms of micropiles under seismic action and offers valuable guidance for the design of slope support.