Site Loader

Crowd detection is a highly focused area for law enforcement, urban engineering and
traffic management. Public places such as shopping centers and airports are monitored using
closed circuit television in order to ensure normal operating conditions. Automated analysis of
crowd activities using surveillance videos is an important issue for communal security during
violence, strike, heavy gathering allows detection of dangerous crowds and where they are
headed. In case of surveillance, group behavior modeling and crowd disaster prevention people
detection and tracking in crowd is a crucial component for wide range of application. Due to
heavy occlusions, view variations and varying density of people as well as the ambiguous
appearance of body the reliable person detection and tracking in crowd becomes a challenging
task. Computer vision based crowd analysis algorithm can be divided into some groups; people
counting, people tracking and crowd behavior analysis, movement analysis.
Person detection and tracking in crowd is a challenging task. Individual object detection
has been improved significantly in recent times but the crowd detection and tracking contains
some challenges. The head density of one person could be similar to another person density.
The density of pedestrians significantly impacts their appearance in a video. For instance, in the
videos with high density of crowds, people often occlude each other and usually few parts of the
body of each individual are visible. On the other hand, the full body or a significant portion of the
body of each pedestrian is visible in videos with low crowd-density. These different appearance
characteristics require tracking methods which suite the density of the crowd 1.This research proposed a system that detect the head region and based on this head
region that can detect people from crowd. The more accurate head detection can lead a good
result for detecting a person in crowd domain. This research focused on gradient feature based
image analysis and found a good accuracy rate of head detection described based on below
Figure 1 and Figure 2 as sample.
Figure 1. Sample Input Image to Detect Crowd Figure 2. Detected Crowd Result
Concentrated on image gradient based people detection. Image gradient basically
contains the directional changes information. This information can be used to track different
objects or regions as well as boundary shape, getting a rough idea of an object location and
other information. By analyzing the regions an assumption of item can be found that it is human
or not in another word we can say that this step is important part selection or interest point
detection. Natural images contain a lot of changes in orientation. So the number of important
part may be large and huge as it is counting based on orientation information. There is some
types of methods is needed to reduce this important part such as Adaboost and others. We
applied different feature extraction technique to detect human on that region or from crowd
place. We analyzed with HOG, SIFT and SURF feature. We used HOG and SIFT combined
feature to test the result.
Assume a section that is a strong candidate for head region means an interest point
that’s may be head or not, is compared with trained support vector machine. Applying manual
annotation technique we have prepared two classes of data, one is positive dataset another is
negative dataset. During dataset preparation we have developed dynamic patch selection and
its size. Supervised SVM is used to train with two dataset. All the candidate regions are tested
with SVM. This test said which one is head or not. We got a marked output that processed with
proposed method.
Next sections are organized as follow. Details of implementation in section three,
preparing dataset with manual annotation in section four, experimental results are shown and
compared with different methods and the next section contains the conclusion.  

Post Author: admin