Convolutional sparse coding (CSC) plays an essential role in many computer vision applications ranging from image compression to deep learning. In this work, we spot the light on a new application where CSC can effectively serve, namely line drawing analysis. The process of drawing a line drawing can be approximated as the sparse spatial localization of a number of typical basic strokes, which in turn can be cast as a non-standard CSC model that considers the line drawing formation process from parametric curves. These curves are learned to optimize the fit between the model and a specific set of line drawings. Parametric representation of sketches is vital in enabling automatic sketch analysis, synthesis and manipulation. A couple of sketch manipulation examples are demonstrated in this work. Consequently, our novel method is expected to provide a reliable and automatic method for parametric sketch description. Through experiments, we empirically validate the convergence of our method to a feasible solution.

Images can exhibit diverse attributes and characteristics, because of variations in both the quantity of each intensity level and their respective positions, histograms display varying distributions. Some images feature symmetric histograms, while others exhibit asymmetry. In image segmentation tasks, traditional mean-based thresholding methods work well with symmetric histograms, relying on Gaussian distribution definitions. However, situations arise where asymmetric distributions must be considered. Threshold-based segmentation entails the partitioning of intensity levels into separate regions determined by the threshold value. Within this category of thresholding methods, Minimum Cross Entropy Thresholding (MCET) stands out as a mean-based thresholding technique with a unique self-contained objective function adaptable to various distributions. In our study, we propose incorporating the Maxwell-Boltzmann Distribution into MCET's objective function. We introduce a specialized model aimed at enhancing efficiency in image segmentation tasks, enabling precise data analysis customized to specific images with histograms skewed to the right. This approach yields improved segmentation results by considering the impact of utilizing the Maxwell-Boltzmann distribution with the right-skewed distribution within MCET's objective function. We have validated our approach, and we conducted a comparative analysis, assessing the performance of our proposed model against relevant studies in the literature. We applied this approach to Sentinel-2 satellite imagery for water body segmentation. The outcomes showcase the effectiveness of our model in segmenting images with right-skewed histograms, substantiated by a variety of performance evaluations.

Hotspot creation is one of the most important modules within virtual environments which helps show the navigators of these environments some information about semantic elements within it and facilitate the navigation between the virtual spaces. In this paper, a system for automatic hotspot proposals and creation in virtual environments is proposed. The system uses computer vision modules to automatically propose hotspot locations in addition to identifying and creating these hotspots with candidate labels. Two main modules used in the system are object detection and scene segmentation. The scene segmentation helps give candidate hotspot areas and provides an overall understanding of the semantics of the virtual environment. The object detection module also uses pretrained deep networks for automatic hotspot creation over these objects. The system helps speed up the hotspot creation process and offers a tool for virtual environment users and creators.

Vanishing point detection is a fundamental step in computer vision that allows 3D scene understanding and autonomous navigation. Classical techniques have significant challenges when trying to understand scenes that are heavily cluttered and images containing multiple perspective cues, leading to poor or unreliable vanishing point determination. We present a Dual RANSAC with Rescue Midpoint-based Multi-Trend Vanishing Point Detection framework, which targets the simultaneous detection and fine-tuning of multiple, globally consistent vanishing points. The proposed framework introduces a novel Midpoint-based Multi-Trend Random Sample Consensus formulation that operates on line segment midpoints to infer dominant directional groups, thereby eliminating noisy or unstable midpoints and stabilizing subsequent vanishing point inference. The main novelty lies in using line segment midpoints to model the orientation variation as a linear regression in the midpoint–orientation space, which helps reduce sensitivity to endpoint instability. Candidate vanishing points are prioritized through inlier-based confidence ranking and subsequently optimized via an MSAC-based arbiter to resolve hypothesis conflicts and minimize geometric error. We evaluate our work against state-of-the-art techniques such as J-Linkage and Conditional Sample Consensus, over two of the current challenging public datasets that comprise the York Urban Dataset and the Toulouse Vanishing Point Dataset. The results show that the proposed framework achieves a recall of up to 95% and an image success rate of almost 84%, outperforming both J-Linkage and Conditional Sample Consensus, especially under tighter angular thresholds. This demonstrates the ability of the proposed framework to provide enhanced stability and localization accuracy.

Convolutional Sparse Coding (CSC) is a well-established image representation model especially suited for image restoration tasks. In this work, we extend the applicability of this model by proposing a supervised approach to convolutional sparse coding, which aims at learning discriminative dictionaries instead of purely reconstructive ones. We incorporate a supervised regularization term into the traditional unsupervised CSC objective to encourage the final dictionary elements to be discriminative. Experimental results show that using supervised convolutional learning results in two key advantages. First, we learn more semantically relevant filters in the dictionary and second, we achieve improved image reconstruction on unseen data.