Deepa Sankar

A new fast fractal modeling approach for the detection of microcalcifications in mammograms.

In this paper, a novel fast method for modeling mammograms by deterministic fractal coding approach to detect the presence of microcalcifications, which are early signs of breast cancer, is presented. The modeled mammogram obtained using fractal encoding method is visually similar to the original image containing microcalcifications, and therefore, when it is taken out from the original mammogram, the presence of microcalcifications can be enhanced. The limitation of fractal image modeling is the tremendous time required for encoding. In the present work, instead of searching for a matching domain in the entire domain pool of the image, three methods based on mean and variance, dynamic range of the image blocks, and mass center features are used. This reduced the encoding time by a factor of 3, 89, and 13, respectively, in the three methods with respect to the conventional fractal image coding method with quad tree partitioning. The mammograms obtained from The Mammographic Image Analysis Society database (ground truth available) gave a total detection score of 87.6%, 87.6%, 90.5%, and 87.6%, for the conventional and the proposed three methods, respectively.

Fractal Modeling of Mammograms Based on Mean and Variance for the Detection of Microcalcifications

In this paper the breast background tissues are modeled using deterministic fractal model based on the mean and variance of the image blocks for detecting the presence of microcalcifications in mammograms are presented. Only those image blocks whose variance difference is between 0.01 and 1 are classified according to their mean value and used in the matching block searching process and therefore the time taken to model the mammograms was considerably reduced to about one third the time required to encode in the conventional fractal encoding scheme. The modeled image will be visually close to the original image and if the difference between the original and the modeled image is taken the presence of microcalcifications can be detected. The method was tested by using the mammograms obtained from MIAS database. The average correlation between the original and the modeled mammograms were obtained as 0.9740 and the average mean square error was found to be 5.939. The results show that the true positive rate is 82% with an average of 0.214 negative clusters per image for 28 mammograms were obtained.

Analysis of mammograms using fractal features

Medical images like mammograms are very difficult to analyze because of their low contrast. Different fractal features are used for analyzing mammograms in this paper. The new fractal feature derived from the modified average image is found to be a better feature for distinguishing between normal, malignant, benign and mammograms with microcalcifications. The study is performed on the mammograms obtained from the online Mammographic Image Analysis Society (MIAS) Digital Mammogram database. The average values of the new normalized fractal feature for normal, mammogram with microcalcifications, benign and malignant tumors are obtained as 0.148, 0.437, 0.3145, and 0.5465 respectively.

Fast Fractal Coding Method for the Detection of Microcalcification in Mammograms

The presence of microcalcifications in mammograms can be considered as an early indication of breast cancer. A fast fractal block coding method to model the mammograms for detecting the presence of microcalcifications is presented in this paper. The conventional fractal image coding method takes enormous amount of time during the fractal block encoding procedure. In the proposed method, the image is divided into shade and non shade blocks based on the dynamic range, and only non shade blocks are encoded using the fractal encoding technique. Since the number of image blocks is considerably reduced in the matching domain search pool, a saving of 97.996% of the encoding time is obtained as compared to the conventional fractal coding method, for modeling mammograms. The above developed mammograms are used for detecting microcalcifications and a diagnostic efficiency of 85.7% is obtained for the 28 mammograms used.

Cascaded H bridge multilevel inverter topologies for PV application: A comparison

Multilevel inverters acts as a promising solution for medium voltage, high power applications due to their modularity and reduced voltage stress across the switches. Cascaded H Bridge Multilevel Inverters (CHB-MLI) are being considered as the best choice for grid connected Photovoltaic (PV) systems since they require several sources on the DC side. By means of MLIs, high quality output with less harmonic distortion is obtained compared to a two level inverter. In this paper, a comparative analysis of four levels of MLIs is presented. Control scheme based on Sinusoidal Pulse Width Modulation (SPWM) is adopted due to its ease of implementation. More number of levels results in reduced THD and nearly sinusoidal output. Simulation is performed using MATLAB/SIMULINK.

Detection of drowsiness based on HOG features and SVM classifiers

This paper presents an accurate method of drowsiness detection for the images obtained using low resolution consumer grade web cameras under normal lighting conditions. The drowsiness detection method uses Haar based cascade classifier for eye tracking and combination of Histogram of oriented gradient (HOG) features combined with Support Vector Machine (SVM) classifier for blink detection. Once the eye blinks are detected then the PERCLOS is calculated from it. If the PERCLOS value is greater than 6 seconds then the person is said to be drowsy. The presented system was validated by comparing the prediction of the system with that of a human rater. The system matched with the human observer with 91.6 % accuracy.

A novel method for eye tracking and blink detection in video frames

This paper presents a novel method for eye tracking and blink detection in the video frames obtained from low resolution consumer grade web cameras. It uses a method involving Haar based cascade classifier for eye tracking and a combination of HOG features with SVM classifier for eye blink detection. The presented method is non intrusive and hence provides a comfortable user interaction. The eye tracking method has an accuracy of 92.3% and the blink detection method has an accuracy of 92.5% when tested using standard databases and a combined accuracy of 86% when tested under real world conditions of a normal room.

A new method for sorting and grading of mangos based on computer vision system

In this paper a new automated method for sorting and grading of mangos based on computer vision algorithms is presented. The application of this system is to replace the existing manual technique of sorting and grading used in India. The system is developed for Alphonso mangos, the premium variety of mango exported from India. The developed system was able to sort the Alphonso mangos with an accuracy of 83.3% and can identify a defective skin up to an min area of 6.093845×10-4 sqcm.

Fractal signatures for the characterization of mammograms

In this paper mammograms are characterized into normal and those with microcalcifications, benign and malignant masses using fractal signatures and distance measures. Here, the gray level intensity surface of the mammogram is assumed to be a blanket of different thickness. Different areas and volumes of this blanket are used to estimate the fractal dimension and hence the fractal signatures and the distance measures. It is found that the fractal signatures of the normal mammograms are much smaller than the cancerous ones. The results show that the differential distance measure could correctly distinguish the normal mammograms from the cancerous mammograms. The mammograms for the study were obtained from the online MIAS database.

A new gray level based method for visual inspection of frying food items

In this paper a novel method for the visual inspection of frying food items using gray level method is presented. The gray levels of the images of the food items are used for recognition of different stages for frying. Mean gray level of 101.9 was obtained as the threshold of complete frying.