Recent developments in video analysis of sports and computer vision techniques have achieved significant improvements to enable a variety of critical operations. To provide enhanced information, such as detailed complex analysis in sports such as soccer, basketball, cricket, and badminton, studies have focused mainly on computer vision techniques employed to carry out different tasks. This paper presents a comprehensive review of sports video analysis for various applications: high-level analysis such as detection and classification of players, tracking players or balls in sports and predicting the trajectories of players or balls, recognizing the team’s strategies, and classifying various events in sports. The paper further discusses published works in a variety of application-specific tasks related to sports and the present researcher’s views regarding them. Since there is a wide research scope in sports for deploying computer vision techniques in various sports, some of the publicly available datasets related to a particular sport have been discussed. This paper reviews detailed discussion on some of the artificial intelligence (AI) applications, GPU-based work-stations and embedded platforms in sports vision. Finally, this review identifies the research directions, probable challenges, and future trends in the area of visual recognition in sports.

In real-world sports video analysis, identity switching caused by inter-object interactions is still a major difficulty for multi-player tracking. Due to similar appearances of players on the same squad, existing methodologies make it difficult to correlate detections and retain identities. In this paper, a novel approach (DeepPlayer-Track) is proposed to track the players and referees, by representing the deep features to retain the tracking identity. To provide identity-coherent trajectories, a sophisticated multi-player tracker is being developed further, encompassing deep features of player and referee identification. The proposed methodology consists of two parts: (i) the You Only Look Once (YOLOv4) can detect and classify players, soccer balls, referees, and background; (ii) Applying a modified deep feature association with a simple online real-time (SORT) tracking model which connects nodes from frame to frame using cosine distance and deep appearance descriptor to correlate the coefficient of the player identity (ID) which improved tracking performance by distinct identities. The proposed model achieved a tracking accuracy of 96% and 60% on MOTA and GMOTA metrics respectively with a detection speed of 23 frames per second (FPS).

Soccer player and ball detection and tracking have emerged as an area of intense interest among many analysts and researchers. This is because it aids coaches in team performance evaluation and decision-making to achieve optimal results. However, existing methodologies have failed to effectively detect and track the ball when it moves at high velocity and also to track players under occlusion conditions. You only look once (YOLOv3) and simple online real-time (SORT)-based soccer ball and player tracking approach is proposed, for accurately classifying the detected objects in soccer video and track them in various challenging situations. The proposed methodology consists of two parts: (i) YOLOv3 can detect and classify the objects (i.e., player, soccer ball, and background) and eliminate the detected objects outside the playfield as background; (ii) tracking is achieved using SORT algorithm which employs a Kalman filtering and bounding box overlap. The proposed model achieves tracking accuracy of 93.7% on multiple object tracking accuracy metrics with a detection speed of 23.7 frames per second (FPS) and a tracking speed of 11.3 FPS.

Event detection and classification in sports is particularly important in the field of sports data analytics. In this era, various AI methods are utilized to detect events in a football (soccer) game. The use of advanced computational techniques in this area can help to achieve higher accuracy in detecting events. This paper proposes a novel architecture to classify events in a football game. The dataset which is used consists of numerous images pertaining to different events in a regular football match. From this dataset the objective is to detect specific and semantically meaningful events such as pass, dribble or shoot, etc. The proposed model consists of a convolutional autoencoder coupled with a Multi-Layer Extreme Learning Machine to classify the given images into defined events. The Convolutional Autoencoder (AE) module is placed before the ELM in the model pipeline as it helps with data compression and dimensionality reduction so as to ensure fast computation. Thereby, a Multi-Layer Extreme Learning machine achieves better performance compared to other existing approaches.

The capacity to comprehend and anticipate the movements of dynamic objects deployed in sports contexts is a need for autonomous systems. A basketball game is an example of one such ambiance, with its dynamic and intricate motions fueled by numerous social interactions. One such crucial part of such systems is to forecast future basketball positions to determine whether or not the ball will hit or misses the basket based on those predictions. This is a difficult task as basketball motion depends on several factors, including the stroke applied by the player on the ball, environmental factors like topology and geometry, etc., and the subject of analyzing the trajectory of basketball is one in which particularly interested. In this paper, it was concentrated on predicting basketball motion trajectories in actual basketball tournaments and the objective is to forecast basketball's future locations based on its past positions. Therefore, this paper investigates whether long short-term memory (LSTM) models can accurately anticipate the motion trajectories of basketballs in light of their performance in sequence prediction challenges. In particular, it was suggested a long-short-term memory network-integrated relation network-based bimodal exponential normal distribution (BEND) process.