Abstract: In order to overcome the problem of unclear regional outline and weak anti-noise ability in image salient region detection, a multi-scales image saliency detection based on block contrast is proposed. This method based on Itti’s model, extracts image features under different scales to express the overall characteristics of images, and then computes local contrast of image block as the saliency value of the image. It uses adaptive threshold value method to extract salient region from the saliency map. Simulation experiment result demonstrates that the method can accurately extract image salient region with clear boundary.

Key words: salient region detection, multiple scales, image block, contrast value

摘要： 针对图像显著区域检测区域轮廓不明确，抗噪能力弱的问题，提出一种基于分块对比的多尺度图像显著区域检测。该方法以Itti模型为基础，在多尺度下提取图像特征以更全面地表现图像的总体特征；以图像块为单位计算图像的局部对比度作为图像的显著值；用自适应阈值法从显著图中提取显著区域。仿真实验结果表明，该方法能够准确地提取图像的显著性区域，使区域具有明确的边界。

关键词: 显著区域检测, 多尺度, 图像分块, 对比显著度