R. I. Minu

Image segmentation using improved JSEG

Multi-class image semantic segmentation (MCISS) is one of the most crucial steps toward many applications such as image editing and content-based image retrieval. Its a very efficient method that include top down and bottom up approaches. In the top down approach model based segmentation is done. Semantic segmentation of image is one which groups the pixels together having common semantic meaning. This is done by applying semantic rules on the image pixels. Semantic texton forest (STF) is used for implementing this approach. In the bottom up approach using JSEG a region based segmentation is performed. To segment an input image, it heuristically groups the pixels in the input image according to their spatial adjacency, boundary continuity etc, and thus have no knowledge about the correspondence between pixels or regions to semantic categories, but will get more accurate boundaries than top down approach. But for some class of images JSEG showing reduced quality segmentation. To solve this FRACTAL JSEG method uses local fractal dimension of pixels as a homogeneity measure. This method showing improved result comparing to JSEG in boundary detection and hence segmentation. Another approach called I-FRAC also showing better results for some class of images where variation of colours is too low. Hence in this work an approach that uses both algorithms based on a selection criteria is proposed. This work is based on the assumption that by improving the bottom up approach using fractal dimension concept segmentation accuracy of MCISS can be improved. Here in the bottom up approach an improved version of JSEG is implemented to focus on how to find out a class specific value for region merging parameter that will increase the accuracy of segmentation.

Wireless Soil Monitoring Sensor for Sprinkler Irrigation Automation System

In a developing country like India, there is an exponential rise in population nutrition requirement. To meet up with both the ends, the agricultural techniques should be perfected for optimal yield and quality. Irrigation and soil property monitoring system using sensors can be automated and operated wirelessly to achieve optimal water supply control and surveillance. The objective of this paper, is to automate the whole wireless sensor network (WSN) system with a control over water pumps and dripper valves. The humidity, temperature and pH sensor’s percepts provide a feedback, to control the water content of the soil. The system has an low-cost and energy reliable ZigBee for sensor data transformation, high-range GPRS system for data storing and analysis, and the whole system is powered by Solar panels which makes it self-sustainable. Customizable options for different crop with different requirements make it a versatile WSN system for automated irrigation based water management.

Multimodal Fuzzy Ontology Creation and Knowledge Information Retrieval

The main objective of this paper is to design an information retrieval system for both text and image data using the concept of ontology. The focus of this paper is to improve information retrieval for sports events using Ontologies. For this purpose, an integration of domain knowledge with images using fuzzy ontology technique was implemented. In this work, the domain of basketball event is considered for creating low-level visual ontology for certain sport event images. The created multimodal ontology definition will provide a wide domain applicability, which allows the user to construct an ontology for any sport event. The domain ontologies are compared with the non-ontological information retrieval system to show the effectiveness of the ontological system.

Improving the conspicuity of lung nodules by use of "Virtual Dual-Energy" radiography

In the field of medical image processing, computer programs have been developed and approved for use in clinical practice that aid radiologists in detecting the abnormalities on radiology exams. In this study, a Computer-aided detection (CADe) scheme with improved sensitivity and specificity is developed. Chest radiograph(CXR) images are used as the input, which is then segmented using Multi segment active shape model (M-ASM). Massive Training Artificial Neural Network(MTANN) is used to suppress the ribs and clavicles as a result of which, Virtual Dual-Energy(VDE) image is developed. In addition, an Hop-Field Neural Network(HNN) is used to improve the rib contrast. Features are extracted from the original image and the VDE image. A nonlinear support vector machine(SVM)classifier was employed for classification of the nodule candidates and a linear discrimination analysis is used to detect the nodules.