Skip to content
Open access

FPGA-Based Bilinear Interpolation Image Scaler with Q8 Fixed-Point Arithmetic and Hardware-Software Co-Verification

Jul 2026 · Computers and artificial intelligence · 0 citations · 10 references

Abstract

Image scaling serves as a critical performance bottleneck in real-time video pipelines powering multi-resolution embedded display systems. The design integrates a dual ping-pong line buffer architecture to deliver four pixels per clock cycle, paired with a digital differential analyzer (DDA) for precise sub-pixel coordinate generation. A four-stage pipelined interpolation core optimized for Xilinx Digital Signal Processing 48 (DSP48) slices achieves one-pixel-per-cycle throughput with only six cycles of end-to-end latency. Hardware-software co-verification conducted on ModelSim SE-64 demonstrates perfect reconstruction of constant-color images and 25.8 dB Peak Signal-to-Noise Ratio (PSNR) for gradient images at 4x4 identity scaling. Resource estimates for a Xilinx Artix-7 XC7A35T indicate the design occupies merely 3.8% of available Look-Up Tables (LUTs) and 6.7% of DSP48 slices, making it exceptionally suitable for low-power embedded applications.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.