摘要
应用计算机数字图象处理技术对木材加工业中的一种典型缺陷──漏节进行了处理，使其比原始图象更清晰，更易于识别。作者在论文中首次提出应用最大信息熵来决定二值化图象的阈值，取得了满意的结果。考虑到实际处理时的实时要求，应用改进后的罗泊特梯度算子来提取漏节的轮廓，使处理时间大大缩短。
A reliable methodn on destructivex ray test of wood defects with computer digital image processing technique was presented in the paper which made the knot hole in wood clear and easily recogniged.It was the first time to use maximum entropy to determine the threshold of binary picture and the result was satisfying.The improved Robert was used in the paper to outline the knot-hole so that the processing was almost on real time.
出处
《东北林业大学学报》
CAS
CSCD
北大核心
1994年第5期69-75,共7页
Journal of Northeast Forestry University
基金
林业部"七.五"攻关项目
关键词
漏节
图象处理
无损检测
木材
Knot-hole
Image processing
Non-destructive test
threshold
