Extrinsic parameter calibration is the foundation and prerequisite for LiDAR and camera data fusion of the autonomous system. This technology is widely used in fields such as autonomous driving, mobile robots, intelligent surveillance, and visual measurement. The learning-based method is one of the targetless calibrating methods in LiDAR and camera calibration. Due to its advantages of fast speed, high accuracy, and robustness under complex conditions, it has gradually been applied in practice from a simple theoretical model in just a few years, becoming an indispensable and important method. This paper systematically summarizes the research and development of this type of method in recent years. According to the principle of calibration parameter estimation, learning-based calibration algorithms are divided into two categories: accurate calibrating estimation and relative calibrating prediction. The evolution routes and algorithm frameworks of these two types of algorithms are elaborated, and the methods used in the algorithms’ steps are summarized. The algorithm mechanism, advantages, limitations, and applicable scenarios are discussed. Finally, we make a summary, pointing out existing research issues and trends for future development.

With the continuous development of multimedia, more and more digital works such as videos are spread, stored, and used in the network. In recent years, digital copyright infringement disputes have occurred frequently. The traditional copyright protection system has some problems, such as difficulty confirming copyright, monitoring infringement, and obtaining evidence for rights protection. To this end, we have designed and implemented a novel video copyright protection scheme based on the blockchain and double watermarking technology. We use the image correlation coefficient method to extract video keyframes. And we combine with Contourlet Transform domain, QR decomposition, and SIFT algorithm to improve the robustness of watermark against geometric attacks on the premise of invisibility. After that, we use Arnold Transformation (Cat Map) based on the Maximum Entropy Threshold Segmentation to encrypt the robust watermark and strengthen the security. Moreover, based on the characteristics of the fragile watermarking, we accurately locate the attacked video’s tamper position and complete the integrity authentication of the watermarked video. In addition, the hash digest of the video watermark and the user ID of the copyright owner is signed by SM2 and uploaded to the blockchain. The user can register the copyright after passing the identity authentication. We conduct tests and security analysis on the blockchain performance of the system, the performance of the commercial cryptography algorithm, and the security of the watermarking system. The experimental results show that the blockchain used in this system conforms to the industry standard, the performance of SM2 and SM3 is better than ECC-256 and SHA-256, and the system security is well guaranteed.

Panoptic segmentation, as a key task in the field of computer vision, holds significant importance in practical applications such as autonomous driving and robot vision. Currently, among deep-learning-based panoptic segmentation methods, query-based methods have received widespread attention. However, existing methods, such as Mask2Former, typically rely on a static query mechanism. This makes it difficult for the model to adapt to changes in the number of instances in different scenes and can lead to instance loss or confusion, thus limiting performance in complex scenes. Furthermore, it is prone to insufficient feature extraction and a loss of global information. To address these problems, this paper proposes a panoptic segmentation method based on dynamic instance queries (PSM-DIQ). PSM-DIQ uses a multi-dimensional attention mechanism to enhance feature extraction, utilizes instance-activation-guided dynamic query generation to improve the ability to distinguish between different instances, and optimizes pixel–query interactions through a dual-path Transformer decoder. Experiments on the Cityscapes and MS COCO datasets show that, based on the ResNet-50 backbone, PSM-DIQ significantly outperforms the Mask2Former baseline, with PQ values improving by 1.8 and 1.7 percentage points, respectively. The experimental results verify the effectiveness of PSM-DIQ in complex scene panoptic segmentation. Finally, this work will be released as an open-source software package on GitHub (v1.0).

Panoptic segmentation, a pivotal research direction in computer vision, unifies pixel-level object and background recognition within a scene, crucial for applications like autonomous driving. However, existing methods, including State-of-the-Art models like Mask2Former, often exhibit limitations such as inadequate adaptation in multi-scale feature fusion and ambiguous boundary segmentation, particularly for small objects in complex scenes. To address these specific challenges, we propose a novel network: PSM-FFEG (Panoptic Segmentation Model with Feature Fusion and Edge Guidance). PSM-FFEG introduces three key components: (1) A dynamic multi-scale feature fusion module enhancing contextual modeling via cascaded convolutions and adaptive attention. (2) An explicit edge guidance module refining boundary features with dedicated supervision. (3) A dual-path Transformer decoder optimizing cross-path feature interaction between pixels and queries. Extensive experiments on the Cityscapes and MS COCO datasets demonstrate that, using a ResNet50 backbone, PSM-FFEG achieves 2.6% and 2.4% absolute improvements in panoptic quality (PQ) over the Mask2Former baseline, respectively. Notably, PSM-FFEG shows significant gains for small objects on Cityscapes, with PQ increasing by 4.3% for traffic lights and 6.8% for motorcycles. These results validate the effectiveness of the proposed modules. To foster further research, our implementation code will be made publicly available.

Valve-controlled hydraulic systems are widely used in various hydraulic equipment, but their asymmetric characteristics are the most critical factor restricting further improvements in system performance. This paper takes large asymmetric complex hydraulic equipment as the industrial background, and proposes a state feedback-based symmetric switching control method to address the complex control strategy and difficult control accuracy caused by input–output asymmetry and the inconsistent response of asymmetric valve-controlled hydraulic systems. A system state space model is established, and the parameterized expression that satisfies the state space switching-based symmetric control law is solved. Feedback and feedforward links based on state space symmetric switching are designed to transform the asymmetric system into a state space symmetric system. And the research results will be applied to the experimental setup of the 300 MN forging hydraulic press control system. Through simulation verification, under asymmetric PID control conditions, due to the influence of the asymmetric characteristics of the system structure, load, and their coupling relationship, the forward response time is shorter than the unloaded response time, and the overshoot is larger than the unloaded response time. The reverse response time is longer than the unloaded response time, and the overshoot is smaller than the unloaded response time. After symmetric control, the forward and reverse dynamic system characteristic curves completely overlap, proving that the system has achieved symmetric transformation; through experimental verification, under asymmetric PID control conditions, when the proportional valve opening remains constant, changes in the load pressure will cause changes in the load speed. For every 1 MPa increase in the load pressure, the load speed will slow down by about 0.0033 m/s. The load speed of the system after symmetrical control replacement will be much less affected by changes in the load pressure. The simulation and experimental results have shown that this method is expected to solve the key problem of inconsistent dynamic characteristics of complex equipment hydraulic systems in both the forward and reverse directions due to asymmetry, and the inability to ensure control accuracy in both directions using symmetric control strategies. This paper has developed a set of control theories and methods applicable to hydraulic systems with complex asymmetry.

The outstanding mechanical properties of nickel-based alloys are predominantly governed by the characteristics of their nano-precipitation. Traditional optimization approaches, however, have focused almost exclusively on average precipitation size, neglecting the non-uniform size distribution induced by the temperature gradient during the actual preparation process. This oversight leads to inaccurate optimization parameters, hindering the reliable design and broader engineering application of these alloys. This study investigates the coupling effect of average size and deviation on the mechanical behavior of nickel-based superalloys by experiment and simulation. The characterization of prepared nickel-based alloys confirms the spatial non-uniform distribution of the precipitation size and calibrates dislocation dynamics (DD) simulation inputs. When the precipitation size exhibits no deviation, conventional strengthening models align with simulation results, confirming the accuracy of the simulations. Under high size deviation, however, significant discrepancies arise between traditional models and simulation outcomes, indicating that conventional models fail to account for the effect of size deviation. This is because size deviation leads to spatial heterogeneity in precipitate distribution, which is difficult to describe using a unified theoretical framework. Machine learning (ML)-augmented analysis of the extensive DD dataset reveals a previously unrecognized synergistic effect governed by precipitation size heterogeneity. Surprisingly, contrary to the conventional preference for a uniform size distribution, we find that an appropriate size deviation increases the number of precipitates that effectively pin dislocations in three-dimensional space, thereby enhancing strength at a constant volume fraction. Our combined experimentally calibrated DD/ML approach identifies an optimal range of size variation that maximizes strengthening potential. These results provide a new strategy and an effective pathway for performance optimization of nickel-based alloys.