Vision Based Umpire Detection method for Event Extraction in Cricket Video based on HOG and Color image Segmentation

As of late, there has been expanded interest in Video Summarization and Automatic Highlights age. One of the most watched and a played sport is cricket, especially in South Asian Countries. In cricket sport video, Umpire has the force for settling on huge choices about functions in the field. In this work, the manually chosen Recorded Video of Some International Twenty-Twenty cricket match video (T20*20) is used as dataset for Umpire identification inside the sport of cricket. The proposed dataset is assessed as a starter help for creating frameworks to consequently produce cricket features. With the growing increase of the utilization of technology in sports, this paper points basically at the Umpire Identification utilizing Area-Of-interest based Histogram of Oriented Gradients (AOI-HOG) Feature Extraction strategy for Human Identification. It expects to give Color Image Segmentation technique utilizing pixel wise color information to support vector machine (SVM) classification for Umpire Identification while protecting the arrangement of highlights. Test proof shows that the proposed technique accomplishes the Accuracy of 94.57%. Proposed one efforts towards successful segmentation results and computational conduct, and lessens the time and expands the norm of color image segmentation