基于尺度极值的灰度图像多值化处理  

MULTIVALUE PROCESSING OF GRAY-LEVEL IMAGE BASED ON SCALE EXTREMES

作　　者：李庆丰[1,2] 战德臣[1,2] 

机构地区：[1]哈尔滨大学计算机科学系 [2]哈尔滨工业大学

出　　处：《黑龙江商学院学报》1997年第4期24-29,共6页Journal of Harbin Commercial University(Natural Sciences Edition)

摘　　要：灰度图像的多值化处理是将一幅具有ｌ个灰度级的图像变成一幅具有ｋ个灰度级（ｋ＜ｌ）的图像，是图像分割的一种有效方法。本文介绍了一种基于直方图峰值点分析—尺度极值分析的灰度图像多值化处理方法，并给出了实验结果及分析。该方法可有效地全自动处理多峰直方图的图像。Multivalue processing of gray-level image is a kind of transformation form and image with l gray level to an image with k gray level in which k is less than l .It is an effective method for segmentation.A new method for multivalue processing based on the selection or varied local extreme points of image′s histogram—a method based on scale extremes is presented in the paper.As for the images with multimodal and flat histograms,experimental results and analysis show that the method is capable of processing both images automatically.

关 键 词：图像分割 灰度图象 多值化处理 图象处理 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]