The significance of the DDoS problem and the increased occurrence and strength of attacks has led to the dawn of numerous prevention mechanisms. IP spoofing is most frequently used in denial-of-service attacks. In such attacks, the goal is to flood the victim with overwhelming amounts of traffic, and the attacker does not care about receiving responses to the attack packets. IP spoofing is one of the basic weaknesses in the Internet Protocol to launch the DDOS attack. Each prevention mechanism has some unique advantages and disadvantages over the others. The existing methods become ineffective due to a large number of filters required and they lack in information about where to place the filter. We propose Ingress filter in Extended Inter Domain Packet Key marking system .This paper comprises of two functional blocks namely, Key marking system and filtering blocks. In the marking block, each source is labeled with a key. The key is changed continuously for a certain period of time to provide secured system and is validated at border routers. In the filtering block, spoofed packets are filtered at the border router using Ingress filter to filter beyond periphery routers. The filter placement algorithm clearly put forwards the conditions under which the filter can operate accurately. The accuracy of the proposed systems is validated using Network Simulator (NS-2).

Image processing is a comprehensively research topic with an elongated history. Segmenting an image is the most challenging and difficult task in image processing and analysis. The principal intricacy met in image segmentation is the ability of techniques to discover semantic objects efficiently from an image without any prior knowledge. One recent work presented connected coherence tree algorithm (CCTA) for image segmentation (with no prior knowledge) which discovered regions of semantic coherence based on neighbor coherence segmentation criteria. It deployed an adaptive spatial scale and a suitable intensity-difference scale to extract several sets of coherent neighboring pixels and maximize the probability of single image content and minimize complex backgrounds. However CCTA segmented images either consists of small, lengthy and slender objects or rigorously ruined by noise, irregular lighting, occlusion, poor illumination, and shadow.
In this paper, we present a Cluster based Semantic Coherent Tree (CBSCT) scheme for image segmentation. CBSCT's initial work is on the semantic connected coherence criteria for the image segregation. Semantic coherent regions are clustered based on Bayesian nearest neighbor search of neighborhood pixels. The segmentation regions are extracted from the images based on the cluster object purity obtained through semantic coherent regions. The clustered image regions are post processed with non linear noise filters. Performance metrics used in the evaluation of CBSCT are semantic coherent pixel size, number of cluster objects, and purity levels of the cluster, segmented coherent region intensity threshold, and quality of segmented images in terms of image clarity with PSNR.

Image segmentation is a vital part of many applications because it makes possible for the information extraction and analysis of image contents. For image segmentation process, many approaches have been proposed earlier. The image segmentation using normal standard methods is well only for simple image contents. In existing approaches image segmentation is done through multi-resolution stochastic level set method (MSLSM), but the topology changes are undesirable and it presented nonparametric topology-constrained segmentation model. To improve the image segmentation more effective, in this work, we plan to do image segmentation using level set method with geodesic active contour to analyze medical image disease diagnosis. With level set segmentation based on geodesic active contour, active cluster objects are segregated. Cluster object formation is done with fuzzification of growing active contours with the previously known contours of diseased image portions. With segment portions new similar regions can be traced out with automatic seeded growing method. Experimentation is conducted with bench mark data sets obtained from UCI repository and proved that the proposed GAC work will be 80% efficiency for image segmentation compared to an existing MSLSM. The parametric evaluations are carried over in terms of segment size and contour growth, Contour objects in the cluster, Similarity ratio of known and unknown contours, Seed size of the detected contours.