STEM与日常科技·英语30篇（3）

30 / 30

JPEG压缩伪影与基于分块编码的固有局限

JPEG divides images into 8×8 pixel blocks before applying discrete cosine transform and quantization.
High compression ratios force aggressive quantization, discarding fine frequency details from each block.
When reconstructed, these simplified blocks often show visible edges where their colors or brightness differ slightly.
These square-shaped distortions are called blocking artifacts—and worsen near sharp transitions like text or hair.
JPEG cannot smooth boundaries between blocks because it processes them independently without overlap.
Low-light photos or heavily compressed web thumbnails reveal these artifacts most clearly under magnification.
Modern formats like WebP and AVIF reduce blocking by using larger transforms or adaptive block sizes.
Still, JPEG remains widely used because its trade-off between size and quality suits everyday sharing well.
Understanding blocking helps photographers choose optimal settings before uploading high-detail images.
It also reminds us that all lossy compression sacrifices some truth for practical convenience.