The challenging issue in the design and deployment of Wireless Sensor Networks (WSNs) is the key management and authentication scheme due to the constraints in the sensor networks. The major constraints in Wireless Sensor Networks are large memory storage, more computational complexity and limited resource. Hence, in order to overcome these constraints and to achieve secure communication between sensor nodes, it is important to establish an efficient key predistribution mechanism. Inspite of the fact that many elegant and clever solutions have been proposed, no practical efficient key predistribution has emerged. The existing key management scheme in WSN using ECC provides a predistribution scheme with bigger key sizes and increased memory overhead. The computational complexity is also high which increases the processing time. The recent progress and research on HECC provides new opportunities to utilize public-key cryptography in Wireless Sensor Networks. The key generation for HECC polynomial using genus-2 curve was performed. The encryption and decryption algorithm for HECC was formulated. The key predistribution using HECC and ECC were implemented in wireless sensor network and simulated using NS2 simulator. The various performance analysis namely delay, throughput and power for both HECC and ECC were performed and the results are shown. It is inferred from the results that the proposed HECC scheme outperforms the existing ECC scheme. Further in this project work, the Blind Signature using HECC and Digital Signature using HECC has been implemented in WSN using NS2. The various performance metrics for both the signature schemes have been obtained and the results were compared.
Nowadays digital camera technology and video processing techniques are increased worldwide. Due to this, the conventional fire detection methods are going to be replaced by computer vision based systems. The computer vision based systems detection has a significant role with surveillance system. Most of the algorithms used in the existing techniques propose spectral, spatial, temporal and other low level features of fire for distinguishing it from other objects in video sequences. This paper proposes a new approach to computational vision-based fire and flame detection by using a fuzzy logic edge detection and motion detection with ANN-SVM classifier as classification tool. The edge detection using fuzzy canny edge detection technique and the motion detection using motion estimation are use for fire and flame detection and ANN-SVM classifier is useful for the final classification. Finally, it decided whether the objects that have changed in that video are flame or not. Therefore, this method detects both smoke and flame effectively and obtain high accuracy by reduce false alarm rate.
In image processing an essential step is image segmentation. The aim of segmentation is to simplify and to change the representation of an image into a form easier to analyse. Many image segmentation methods are available but most of these methods are not suitable for thermal image and they need prior knowledge. In order to overcome these obstacles, a new thermal image segmentation methodis developed using an unsupervised artificial neural network method called Kohonen's self-organizing map and a threshold technique. Kohonen's self-organizing map is used to organize the pixels according to Gary level values of multiple bands into groups then a threshold technique is used to cluster the image into dislocate zone, this mode is TSOM.