A novel image classification and image retrieval using saliency driven multiscale nonlinear diffusion filtering and linear distance coding

Feature extraction-coding-pooling framework generates robust and discriminative image representation which improves the performance of image classification. Image classification and image retrieval is carried out using saliency driven multi-scale nonlinear diffusion filtering with linear distance coding (LDC). The saliency driven multi-scale nonlinear diffusion filtering generates images at small, mid and high scale. Information from these scales produces robust image classification using LDC. The proposed image retrieval system is the content based image retrieval. The class manifolds generated by this classification system are given as input to image retrieval system which finally retrieves all images relevant to supplied image.