Digital image processing techniques for estimating power released from the corona discharges

Digital image processing is being widely used in all fields of science and engineering, like the detection of cancer cells, identification of the soil properties etc. Application of image processing to high voltage phenomenon has not been well explored. The present work attempts to identify the possible application of image processing techniques to obtain the image parameters which can be used as indicators to estimate the corona power. A color thresholding based image segmentation method is employed on the digital images of corona and thereby effective spread of corona plasma has been extracted. The variations in the parameters: average intensity, area of the spread, energy, entropy and luminosity content in the processed images are studied in relation with the measured power. It is observed that the parameters extracted from the corona images share a unique relation with the measured corona power, and can be used for the prediction analysis.