An effective median filter for salt & pepper impulse range signals. An additive noise process may corrupt these noise removal is presented. This computationally efficient filtering digital images in both the acquisition and transmission stages. technique is implemented by a two pass algorithm: In the first Application-specific image filtering algorithms are needed to pass, identification of corrupted pixels that are to be filtered are simultaneously remove the effects of the corruptive process perfectly detected into a flag image using a variable sized detection window approach; In the second pass, using the and preserve important features of the images. Impulse noise detected flag image, the pixels to be modified are identified and removal in image processing often involves the removal of corrected by a more suitable median. Experimental results show these salt and pepper noises from images which is a very that the proposed algorithm performs far more superior than important pre-processing step for most other subsequent many of the median filtering techniques reported in terms of processing tasks such as edge detection, segmentation and retaining the fidelity of the image highly corrupted by impulse classification. In this area, early advances were dominated by noises even to the tune of ninety percent impulse noise. The linear filtering. They have had enormous impact on the proposed algorithm is free from patchy effects, does not extend lmelterf They techad formous stationary black or white blocks in the image as has been found in many development of various techniques for processing stationary other adaptive median based techniques and is very effective in and non-stationary signals. However, there are a large number cases when images are corrupted with large percentage of impulse of situations where linear filtering approach performs poorly. noises. This algorithm works very well for images with lower The limitation is the inability to simultaneously eradicate noise percentage of impulse noises also.
Human face detection techniques play an important role in applications like face recognition, video surveillance, human computer interface, face image database management, and querying image databases. Using color information in images is one of the various possible techniques usedfor face detection. This paper proposes a novel techniquefor detecting faces in color images using an adaptive threshold and template matching techniques. The goal of the technique is to segment the skin regions from the non-skin regions. Experimental results demonstrate successful face detection over a wide range offacial variations in color, position, scale, orientation, 3D pose, and expression in images from several photo collections (both indoors and outdoors). This method is quite practical andfaster when compared to neural networks and other techniques especially suited well for multimedia applications on the web.
World wide web is a huge information source, broadly used for learning now-a-days due to flexibility of time, sharing of learning resources and infrastructure etc., Most of web based learning system lacks expert-learner interaction, assessment of user activities and learners are getting drowned by huge number of web pages in the learning web site and they find difficulties in choosing suitable materials relevant to their interest. This work attempts to engage e-learners at an early stage of learning by providing navigation recommendations. E-learning personalization is done by mining the web usage data like recent browsing histories of learners of similar interest. The proposed method uses upper approximation based rough set clustering and dynamic all k th order association rule mining using apriori for personalizing e-learners by providing learning shortcuts. The essence of combing association rule and clustering is that, using clustered access patterns can reduce the data set size for association rule mining task, and improves the recommendation accuracy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.