A key task for customs workers is the interception of hazardous, illegal and counterfeit items in order to protect the health and safety of citizens. However, it is estimated that only a small fraction of cargo is inspected and an even smaller fraction of trafficked goods are detected. Today, the most widely used technology for scanning vehicles, ranging from vans and trucks to railcars, is $\gamma$ ray and X-ray radiography. New technologies are required to overcome current technological shortcomings, such as the inability to detect the target material composition, the usage of harmful ionising radiation sources and the resultant low throughput. Cosmic ray tomography (CRT) is a promising technology for cargo screening. Cosmic ray muons have average energies of around 10,000 times larger than a typical X-ray and therefore can penetrate relatively large and dense materials. By analysing muon scattering, it is possible to identify materials hidden inside shielding that is too thick or deep for other imaging methods. CRT is also completely passive, exploiting naturally occurring secondary cosmic radiation, and is therefore safe for humans and animals. Contrary to conventional X-ray- or $\gamma$-ray-based imaging techniques, CRT also allows material differentiation and anomaly localisation within the cargo or vehicle through the provision of 3D images. This article reviews the current state-of-the-art technology in CRT, critically assessing the strengths and weaknesses of the method, and suggesting further directions for development.

Camera systems support the rapid assessment of ship traffic at ports, allowing for a better perspective of the maritime situation. However, optimal ship monitoring requires a level of automation that allows personnel to keep track of relevant variables in the maritime situation in an understandable and visualisable format. It therefore becomes important to have real-time recognition of ships present at the infrastructure, with their class and geographic position presented to the maritime situational awareness operator. This work presents a novel dataset, ShipSG, for the segmentation and georeferencing of ships in maritime monitoring scenes with a static oblique view. Moreover, an exploration of four instance segmentation methods, with a focus on robust (Mask-RCNN, DetectoRS) and real-time performances (YOLACT, Centermask-Lite) and their generalisation to other existing maritime datasets, is shown. Lastly, a method for georeferencing ship masks is proposed. This includes an automatic calculation of the pixel of the segmented ship to be georeferenced and the use of a homography to transform this pixel to geographic coordinates. DetectoRS provided the highest ship segmentation mAP of 0.747. The fastest segmentation method was Centermask-Lite, with 40.96 FPS. The accuracy of our georeferencing method was (22 ± 10) m for ships detected within a 400 m range, and (53 ± 24) m for ships over 400 m away from the camera.

Assessing the security status of maritime infrastructures is a key factor for maritime safety and security. Facilities such as ports and harbors are highly active traffic zones with many different agents and infrastructures present, like containers, trucks or vessels. Conveying security-related information in a concise and easily understandable format can support the decision-making process of stakeholders, such as port authorities, law enforcement agencies and emergency services. In this work, we propose a novel real-time 3D reconstruction framework for enhancing maritime situational awareness pictures by joining temporal 2D video data into a single consistent display. We introduce and verify a pipeline prototype for dynamic 3D reconstruction of maritime objects using a static observer and stereoscopic cameras on an GPU-accelerated embedded device. A simulated dataset of a harbor basin was created and used for real-time processing. Usage of a simulated setup allowed verification against synthetic ground-truth data. The presented pipeline runs entirely on a remote, low-power embedded system with $$\sim $$∼ 6 Hz. A Nvidia Jetson Xavier AGX module was used, featuring 512 CUDA-cores, 16 GB memory and an ARMv8 64-bit octa-core CPU.

Cosmic ray tomography is an emerging imaging technique utilizing an ambient source of radiation. One common tomography method is based on the measurement of muons scattered by the examined objects, which allows the reconstruction and discrimination of materials with different properties. From the interaction of air shower particles induced through cosmic rays with the material to be scanned, secondary particles, predominantly photons, neutrons and electrons, can be produced, which carry complementary information about the objects and their materials. However, this information is currently not fully exploited or only studied in coincidence with the incoming air shower particles. Therefore, this work presents a novel approach utilizing only the information from secondary particles to reconstruct and discriminate objects made out of a variety of materials. It also includes a detailed analysis of the kinematics of secondary particles and their dependency on material characteristics. In addition, a reconstruction algorithm to produce 3D maps of the examined volume from the measurement of secondary particles is introduced. This results in a successful reconstruction and differentiation of objects in various geometrical compositions.

Particle simulation software is essential for the development and validation of theoretical prototypes in a myriad of physical appplications. The Geant4 simulation toolkit provides precise particle transport and matter interaction simulations in a controlled setting. The interpretation of simulated results is intimately linked to the level of detail represented in the simulation itself, including the ability of the simulation framework to create detailed geometries. Yet, rendering such highly detailed geometries is a daunting task in Geant4, where high-variance scenes must often be manually coded. Potential geometrical errors increase the time required for this coding procedure, and thus in-depth knowledge of the underlying simulation engine in Geant4 is needed.This research proposes Blender2Geant4 (B2G4), a novel framework which transplants 3D scenes from Blender into Geant4 for synthetic data generation. This was achieved by synergizing the descriptive 3D modelling tools in Blender with the simulation capabilities of Geant4. The ability to import, arrange, and manipulate 3D objects in Blender permits the creation of highly detailed and varied scenes: users can easily create sophisticated geometries by using drag-and-drop placement, shape variance randomization, and intuitive material assignments. The suite of functionalities defined in B2G4 lexically translates geometry facets from Blender into a readable format for Geant4, enabling the automated export and import of scenes with minimal manual input. Hence, the B2G4 framework enables automated mass production of corrected 3D scene variations with precise annotations and geometrical consistency checks. The B2G4 framework is designed as an all-purpose geometry creation interface to reduce the complexity of simulation scenes in Geant4 for a variety of physical applications. The applicability of B2G4 in a muon tomography setup is highlighted with a set of procedurally generated 3D scenes.

Muon-scattering tomography (MST) utilizes the deflection of cosmic-ray muons to non-invasively reconstruct the three-dimensional internal structure and material composition of concealed objects, such as those in maritime cargo. Yet, the high dimensionality of reconstructed MST volumes and sparsity of muon hits hinder reliable material discrimination and structural interpretation. We present an unsupervised workflow that visualizes learned data embeddings for material identification. The pipeline couples the Blender-to-Geant4 simulation framework, enabling the rapid prototyping of complex 3D scenes with a standard and widely adopted MST reconstruction algorithm, the Point of Closest Approach (PoCA), to reconstruct the scenes. A structured muon-data sampling grid, termed pillars, feeds an exploratory embedding technique that reveals discriminative material patterns in the reconstructed outputs. Experimental results demonstrate that the proposed approach mitigates key machine-learning challenges in MST; at the same time, they reveal the intrinsic limitations of PoCA estimates for mainstream material classification with machine-learning approaches, and we introduce corrections that enhance visualization and enable data-driven analysis in practical MST deployments.