V - Channel magnification enabled by hybrid optimization algorithm: Enhancement of video super resolution

Area of Science:

• Computer Vision
• Image Processing
• Artificial Intelligence

Background:

• Television super-resolution is a challenging research area.
• Blur motion and computational limitations hinder video enhancement.
• Existing methods struggle with achieving high-resolution video quality.

Purpose of the Study:

• To present a novel smart super-resolution (SR) framework for camera shots.
• To enhance video frames by improving pixel intensity and detail.
• To address the limitations of current video enhancement techniques.

Main Methods:

• RGB frames converted to HSV, with enhancement applied to the V-channel.
• Utilized motion estimation, cubic spline interpolation, and deblurring/sharpening.
• Introduced a hybrid Lion with Particle Swarm Velocity Update (LPSO-VU) algorithm for parameter optimization.

Main Results:

• The LPSO-VU model demonstrated superior performance in video frame enhancement.
• Achieved significant percentage improvements over traditional models like PSO, GWO, WOA, ROA, MF-ROA, and LA.
• Validated superiority using BRISQUE, SDME, and ESSIM metrics.

Conclusions:

• The proposed LPSO-VU framework offers a significant advancement in video super-resolution.
• The hybrid optimization approach effectively overcomes previous enhancement limitations.
• The method provides a robust solution for generating high-resolution video content.