Object Detection From Cluttered Image
The aim of this project is to segment an object from a texture cluttered image. The segmentation is achieved by extracting local information of image and embedding it with active contour model based on region. Images with in-homogenous intensity can be segmented using this model by extracting the local information of image. Segmentation of images with non-uniform intensity can be achieved using the proposed ACM model based on autocorrelation function, which is capable of detecting small objects against a cluttered background. Image features are calculated using a combination of short-term autocorrelations (STA) computed from the image pixels to represent region information. The obtained features are exploited to define an energy function for the localized region-based active contour model called normalized accumulated short-term autocorrelation (NASTA). Minimizing this energy function, can accurately detect small objects in images containing cluttered and textured backgrounds This project constructs local image fitting (LIF) energy functional using local image information. LIF energy functional is defined as an impediment of the differences between the original image and the fitting image. In addition, Gaussian Kernel filtering stabilizes the level set function in each iteration. Another advantage of this proposed model is the elimination of re-initialization. Moreover, the proposed method provides high robustness against random noise and can precisely locate small objects in noisy backgrounds, difficult to be detected with naked eye.