Improved Adaptive Segmenting Algorithm for Steel Bar Image
In order to enhance segmentation result,the end image of bundled steel bar under complicated condition is corrected non-uniform illumination and filtered noise firstly,and a new threshold searching scope and a new threshold criterion for finding the best threshold are put forward to improve maximum between-cluster variance algorithm.Then the end image is segmented through the new method. Finally,it is simulated by Matlab.The experiment results show that the new algorithm not only shortens operation time,but also effectively reduces the interference of uniform illumination and noise. It highlights the details inside the image and contributes to subsequently steel bar identification and counting.
end image of steel barimage enhancementadaptive threshold
NETL
NSTL
万方数据
维普
