Research on the Application of Image Processing and Image Recognition Algorithms in Digital Media in Content Editing and Production

Abstract

Image processing and image recognition algorithms are essential to content generation and editing in the digital age. They provide creative ways to improve visual quality, streamline processes, and customize user interfaces. In-depth methods for incorporating cutting-edge image processing and image recognition algorithms into digital media content production and editing workflows are covered in this study. Methods: To maximize visual quality, the research first focuses on pre-processing and improving digital images using the median filtering technique. The Histogram of Oriented Gradients (HOG) technique is applied to locate and identify the objects in images, that enabling customized and interactive content modification. After that, the images are segmented using the watershed approach, and the precise classification of the images is achieved by applying the Mayfly Optimized Spatial Graph Recurrent Neural Network (MOSGRNN), which improves content organization and retrieval. Results: The results of the experiments demonstrate that the suggested strategy performs well in all aspects of image processing in terms of accuracy (94.91%), recall (92.70%), recognition speed (44 FPS), and f1-score (93.7%). Conclusion: Content production, editing, and delivery across a variety of platforms might be significantly transformed by research on image processing and image recognition algorithms in digital media.