中文摘要：

在阈值分割法中，Otsu算法是图像阈值化分割法的经典算法，在图像处理中有广泛的应用；但此算法存在耗时长、分割精度低等缺点。在对Otsu阈值分割法进行分析的基础上，提出一种改进的Otsu算法。该算法采用缩小可取的灰度值范围的方式减少计算量，在小范围内应用Otsu算法计算最终分割阈值。通过Matlab仿真，所得结果对比表明，改进的Otsu算法在图像分割时间和效果上均优于原Otus算法。

英文摘要：

The Otsu algorithm is a classical algorithm of image threshold segmentation method,and is widely used in imageprocessing. However,this algorithm has the disadvantages of long time consuming,low segmentation precision,etc. On the ba?sis of the analysis of the Otsu threshold segmentation method,an improved Otsu algorithm is presented. The algorithm can re?duce the calculated quantity by narrowing the range of the available grayscale value,and then calculate the final segmentationthreshold in a small area. The contrast results of Matlab simulation show that the improved Otsu algorithm is superior to the origi?nal Otsu algorithm in the aspects of image segmentation time and effect.