یک نظرسنجی در مورد رویکردهای تشخیص و ردیابی شیء مبتنی بر محاسبات نرم
Abstract
1- Introduction
2- Motivation and contribution
3- Neural network based object detection and tracking approaches
4- Fuzzy logic based approaches
5- Evolutionary algorithm based approaches
6- Hybrid approaches
7- Other trends in object detection and tracking
8- Datasets
9- Challenges in object detection and tracking in videos
10- Analysis and inferences
11- Conclusions and summarizations
References
Abstract
In recent years, analysis and interpretation of video sequences to detect and track objects of interest had become an active research field in computer vision and image processing. Detection and tracking includes extraction of moving object from frames and continuous tracking it thereafter forming persistent object trajectories over time. There are some really smart techniques proposed by researchers for efficient and robust detection or tracking of objects in videos. A comprehensive coverage of such innovative techniques for which solutions have been motivated by theories of soft computing approaches is proposed. The main objective of this research investigation is to study and highlight efforts of researchers who had conducted some brilliant work on soft computing based detection and tracking approaches in video sequence. The study is novel as it traces rise of soft computing methods in field of object detection and tracking in videos which has been neglected over the years. The survey is compilation of studies on neural network, deep learning, fuzzy logic, evolutionary algorithms, hybrid and recent innovative approaches that have been applied to field of detection and tracking. The paper also highlights benchmark datasets available to researchers for experimentation and validation of their own algorithms. Major research challenges in the field of detection and tracking along with some recommendations are also provided. The paper provides number of analyses to guide future directions of research and advocates for more applications of soft computing approaches for object detection and tracking approaches in videos. The paper is targeted at young researchers who will like to see it as platform for introduction to a mature and relatively complex field. The study will be helpful in appropriate use of an existing method for systematically designing a new approach or improving performance of existing approaches.
Introduction
The exponential growth in hardware facilities like cameras, processingmachines,mobilephoneshave ledto anexplosionof studies in automated video analysis for object detection and tracking. It is one of the hottest topics of research in computer vision and image processing. Object detection and tracking in video sequence is the key technology in the development of various video analysis applications that tires to detect and track objects over a sequence of images by replacing old traditional methods of monitoring cameras by human operators. The proposed solutions range from low cost handheld devices or cameras to high cost sophisticated and proprietary solutions. Object detection is the process of locating the occurrence of object using number of techniques like background subtraction, feature extraction, statistical methods etc. Fig. 1 shows some example of object detection in videos. Colored boxes on the figures are used to highlight the detected object on a frame. A number of traditional approaches like background detection, frame difference, Gaussian mixture modeling, optical flow based techniques are commonly used to detect objects. Object tracking, on the other hand is process of monitoring an object or multiple objects using a camera over time. Tracking deals with assigning labels to object being tracked and plotting suitable trajectories to specify object motion or deviations. Addition object specific information like area, shape size, orientation can also be extracted by object tracking method. Tracking methods work by detecting the object when it appears for the first time in a frame and predict its trajectories. Such detection-based algorithms estimate the object location in every frame independently. These require an offline training stage and cannot be applied to unknown objects. There is another class of methods where detectors and trackers work simultaneously in which tracker can provide weakly labeled training data for a detector and can help to improve its performance using suitable learning mechanisms. Such trackers can work in online as well offline mode.