Semantic Concept Mining Based on Hierarchical Event Detection for Soccer Video Indexing.

In this paper, we present a novel automated indexing and semantic labeling for broadcast soccer video sequences. The proposed method automatically extracts silent events from the video and classifies each event sequence into a concept by sequential association mining. The paper makes three new contributions in multimodal sports video indexing and summarization. First, we propose a novel hierarchical framework for soccer (football) video event sequence detection and classification. Unlike most existing video classification approaches, which focus on shot detection followed by shot-clustering for classification, the proposed scheme perform a top-down video scene classification which avoids shot clustering. This improves the classification accuracy and also maintains the temporal order of shots. Second, we compute the association for the events of each excitement clip using a priori mining algorithm. We pro- pose a novel sequential association distance to classify the association of the excitement clip into semantic concepts. For soccer video, we have considered goal scored by team-A, goal scored by team-B, goal saved by team-A, goal saved by team-B as semantic concepts. Third, the extracted excitement clips with semantic concept label helps us to summarize many hours of video to collection of soccer highlights such as goals, saves, corner kicks, etc. We show promising results, with correctly indexed soccer scenes, enabling structural and temporal analysis, such as video retrieval, highlight extraction, and video skimming.

Flux Tensor Constrained Geodesic Active Contours with Sensor Fusion for Persistent Object Tracking.

This paper makes new contributions in motion detection, object segmentation and trajectory estimation to create a successful object tracking system. A new efficient motion detection algorithm referred to as the flux tensor is used to detect moving objects in infrared video without requiring background modeling or contour extraction. The flux tensor-based motion detector when applied to infrared video is more accurate than thresholding "hot-spots", and is insensitive to shadows as well as illumination changes in the visible channel. In real world monitoring tasks fusing scene information from multiple sensors and sources is a useful core mechanism to deal with complex scenes, lighting conditions and environmental variables. The object segmentation algorithm uses level set-based geodesic active contour evolution that incorporates the fusion of visible color and infrared edge informations in a novel manner. Touching or overlapping objects are further refined during the segmentation process using an appropriate shape-based model. Multiple object tracking using correspondence graphs is extended to handle groups of objects and occlusion events by Kalman filter-based cluster trajectory analysis and watershed segmentation. The proposed object tracking algorithm was successfully tested on several difficult outdoor multispectral videos from stationary sensors and is not confounded by shadows or illumination variations.