Machine Graphics & Vision https://mgv.sggw.edu.pl/ <p><strong><em>Machine GRAPHICS &amp; VISION</em></strong> is a refereed international journal, published quarterly by the <a href="https://iit.sggw.edu.pl/?lang=en" target="_blank" rel="noopener">Institute of Information Technology</a> of the <a href="https://www.sggw.edu.pl/en/" target="_blank" rel="noopener">Warsaw University of Life Sciences</a> – <a href="https://www.sggw.edu.pl/en/" target="_blank" rel="noopener">SGGW</a>, in cooperation with the <a href="https://tpo.org.pl/" target="_blank" rel="noopener">Association for Image Processing</a>, Poland – <a href="https://tpo.org.pl/" target="_blank" rel="noopener">TPO</a>.</p> <p><strong><em>MG&amp;V</em></strong> has been published since 1992.</p> <p><strong><em>Machine GRAPHICS &amp; VISION</em></strong> provides a scientific exchange forum and an authoritative source of information in the field of, in general, pictorial information exchange between computers and their environment, including applications of visual and graphical computer systems (<a href="https://czasopisma.sggw.edu.pl/index.php/mgv/about">more</a>).</p> Szkoła Główna Gospodarstwa Wiejskiego w Warszawie en-US Machine Graphics & Vision 1230-0535 Optimization of VR human-computer game interaction based on improved PIFPAF algorithm and binocular vision https://mgv.sggw.edu.pl/article/view/10316 <p>To make virtual reality human-computer games more accurate and provide users with an immersive gaming experience, the study combines the method of improved part intensity field and part association field (PIFPAF) with binocular vision to optimize the interaction of VR human-computer games. The experimental results indicated that the PIFPAF algorithms performed relatively well with number of errors and target keypoint correlation of 0.22 and 0.97, respectively. In terms of processing speed, the algorithm performed faster in both 640×480 and 320×240 resolutions, with 13 fps and 19 fps, respectively. Among the five predefined gestures, the ʻʻpointingʼʼ gesture was recognized correctly the largest number of times in 30 test sessions, with 29 successful identifications. In contrast, the ʻʻclenched fistʼʼ gesture had the fewest correct recognitions, totaling 26. The success of the suggested approach is confirmed by the experimental findings, which show that the optimized human-computer interaction system has high accuracy and processing speed. This study offers a fresh approach to the advancement of human-computer interaction technology and encourages technological integration innovation in the realm of virtual reality human-computer gaming.</p> Hong Zhu Bo Chen Copyright (c) 2025 Machine Graphics & Vision 2025-07-26 2025-07-26 34 3 3 30 10.22630/MGV.2025.34.3.1 Advancing plant disease detection: A comparative analysis of deep learning and hybrid machine learning models https://mgv.sggw.edu.pl/article/view/10329 <p>Detecting plant diseases is essential for precision agriculture, as it enhances crop production and ensures the security of the food supply. We adopted two methods for this research: a method based on deep learning, through Convolutional Neural Networks (CNN), and a hybrid model using classical machine learning. The dataset comprised images of plant leaves from Kirtan village in Hisar, Haryana, which were annotated by plant pathologists. The CNN model, which autonomously extracts hierarchical spatial features, achieved an accuracy of 97.57%, making it ideal for large datasets. Conversely, the Hybrid model utilizing handcrafted GLCM and LBP features and SVM classifiers achieved 91.73% accuracy while providing interpretability and computational efficiency in resource limited setups. The performance of the models was measured in terms of accuracy, precision, recall and F1-score. Applications range from on-line monitoring with drones to diagnostic equipment for the farmer.</p> Rajesh Kumar Vikram Singh Copyright (c) 2025 Machine Graphics & Vision 2025-09-24 2025-09-24 34 3 57 75 10.22630/MGV.2025.34.3.3 Fine-tuning stable diffusion for generating 2D floor plans using prompt templates https://mgv.sggw.edu.pl/article/view/10294 <p>Automated generation of 2D floor plans is crucial for architectural design, requiring models to balance precision and adaptability to user-defined specifications. Diffusion models, like Stable Diffusion, excel at generating high-quality images but lack an intrinsic understanding of structured layouts such as floor plans. Conversely, Graph Neural Networks (GNNs) are adept at encoding relational data, representing floor plan objects as nodes and their connections as edges, but they are not generative or capable of processing textual inputs. In this work, we fine-tune Stable Diffusion 1.5 on a custom dataset of floor plans, leveraging structured prompt templates to constrain the model's creativity and guide it toward generating concise, error-tolerant outputs. This research suggests integrating the generative capabilities of diffusion models with the representational strengths of GNNs to overcome inherent challenges in diffusion models, like their inability to explicitly encode spatial relationships. This integration could expand the capabilities of these models, empowering them to comprehend and produce structured layouts more effectively. While computational constraints limited our exploration of this hybrid architecture, our results demonstrate that prompt engineering and dataset preprocessing significantly improve the output quality. This study highlights the potential for generative models in architectural tasks and lays the groundwork for integrating logical reasoning into diffusion-based architectures.</p> Ahmed Mostafa Omar Amir Ali M. Mohamed Marwa O. Al Enany Copyright (c) 2025 Machine Graphics & Vision 2025-09-30 2025-09-30 34 3 77 95 10.22630/MGV.2025.34.3.4 A review: Machine learning techniques of brain tumor classification and segmentation https://mgv.sggw.edu.pl/article/view/10355 <p>Classifying brain tumors in magnetic resonance images (MRI) is a critical endeavor in medical image processing, given the challenging nature of automated tumor recognition. The variability and complexity in the location, size, shape, and texture of these lesions, coupled with the intensity similarities between brain lesions and normal tissues, pose significant hurdles. This study focuses on the importance of brain tumor detection and its challenges within the context of medical image processing. Presently, researchers have devised various interventions aimed at developing models for brain tumor classification to mitigate human involvement. However, there are limitations on time and cost for this task, as well as some other challenges that can identify tumor tissues. This study reviews many publications that classify brain tumors. Mostly employed supervised machine learning algorithms like support vector machine (SVM), random forest (RF), Gaussian Naive Bayes (GNB), k-Nearest Neighbors (K-NN), and k-means and some researchers employed convolutional neural network methods, transfer learning, deep learning, and ensemble learning. Every classification algorithm aims to provide an accurate and effective system, allowing for the fastest and most precise tumor detection possible. Usually, a pre-processing approach is employed to assess the system's accuracy; other techniques, such as the Gabor discrete wavelet transform (DWT), Local Binary Pattern (LBP), Gray Level Co-occurrence Matrix (GLCM), Principal Component Analysis (PCA), Scale-Invariant Feature Transform (SIFT) and the descriptor histogram of oriented gradients (HOG). In this study, we examine prior research on feature extraction techniques, discussing various classification methods and highlighting their respective advantages, providing statistical analysis on their performance.</p> Iliass Zine-dine Jamal Riffi Khalid El Fazazy Ismail El Batteoui Mohamed Adnane Mahraz Hamid Tairi Copyright (c) 2025 Machine Graphics & Vision 2025-09-01 2025-09-01 34 3 31 55 10.22630/MGV.2025.34.3.2