MultiClass Object Classification in Video Surveillance Systems - Experimental Study

Abstract

There is a growing demand in automated public safety systems for detecting unauthorized vehicle parking, intrusions, un-intended baggage, etc. Object detection and recognition significantly impact these applications. Object detection and recognition are challenging problems in this context, since the purpose of the surveillance videos is to capture a wide landscape of the scene; resulting in small, low-resolution and occluded images for objects. In this paper, we present an experimental study on geometric and appearance features (∈ R ≈25000 ) for outdoor video surveillance systems. We also studied the classification performance under two dimensionality reduction techniques (i.e. PCA and Entropy-Based feature Selection). As a result, we built an experimental framework for an object classification system for surveillance videos with different configurations.

Lost and found: Identifying objects in long-term surveillance videos

“…More recently, Elhoseiny et al [14] attempted to classify objects detected in outdoor videos from the VIRAT surveillance dataset into five distinct classes (Human, Car, Vehicle, Objects, and Bicycle) using different types of feature extraction techniques, similar to those from [13]. They reported the weaknesses of appearance-based features such as HOG, and how it can become very effective when appropriately combined with geometric features such as luminance symmetry, central moments and ART moments.…”

“…Though the VIRAT dataset used in [14] is already quite extensive from the aspect of scene and view variability, it does not deal with scenes that are collected over a long period of time, which will present a different challenge altogether. Moreover, VIRAT videos were captured at highdefinition resolutions, which is not the case in many realworld scenarios.…”

Contextual Combination of Appearance and Motion for Intersection Videos with Vehicles and Pedestrians

“…An appropriate object detection and classification method in a video surveillance system should be able to deal with challenging problems like different environment situations, occlusions and low resolution images [3]. Object detection and classification at intersections adds another challenging problem.…”

“…Most traditional object recognition techniques designed for surveillance use motion cues which are not appropriate for intersections. The traditional surveillance issues are addressed in [3] where motion is used to segment moving objects. Different feature extraction techniques like HOG and PCA with SVM are applied for object classification.…”

A Study on Classification for Static and Moving Object in Video Surveillance System

“…to provide security in public area and searching of video. There are various challenges in classification of moving object, due to various reasons: (1) Environment conditions that are not controlled such as fog, rain, and lighting (2) Appearance details of moving objects that is not complete due to ambiguity problem, (3) distance between the moving object and camera is large, (4) video frames contains low resolution quality [18]. Now day's video is becoming an important source for information record.…”