An Efficient Template matching algorithm for Lung Cancer Detection using Multi Resolution Histogram based image segmentation

Various lung cancer detection procedures have been discussed earlier with image segmentation but struggles with accuracy and false positive results. We propose a template matching algorithm for LCD using multi level histogram to segment the pixels of the Lung image to increase the efficiency and accuracy with low time complexity. The proposed method has following steps: Preprocessing, Histogram generation, Segmentation, Template matching, LC identification. The input image applied with preprocessing techniques to improve the quality of image and performed histogram equalization. At the next stage segmentation is performed to group and identify the similar pixels which helps us to find the edges of the left, right lungs. The segmented features are used to extract the features of the lung image. We use template matching technique to scale the portion of the lung and the lung region is extracted. The proposed method maintains the lung image data set which is trained one and using the training set the test image portions are matched to extract the lung portions.