全局低秩显著性检测算法首先根据自然图像前景目标和背景亮度、颜色的差异性重构出图像前景显著目标；然后利用低秩分解对图像中的非显著性区域进行抑制。(The global low-rank saliency detection algorithm first reconstructs the image foreground salient targets based on the difference between the natural image foreground target and the background luminance and color; then, the low-rank decomposition is used to suppress the non-significant regions in the image.)