Page 224 - Proceeding of Atrans Young Researcher's Forum 2019_Neat

“Transportation for A Better Life:
Smart Mobility for Now and Then”

23 August 2019, Bangkok, Thailand

suffering Jesus fallen under the weight of the cross points over a specific length of time was determined
(College of the Holy Cross, 2018). The Black to obtain his or her velocity. The point-process-
Nazarene statue is brought out of the church for based representation described the movement of a
public adoration, in which it is carried along a 6.9 pedestrian as a “go” and the as “stop” when the
km route from Rizal (Luneta) Park to the minor pedestrian was found to be stationary. Pedestrians’
basilica of Quiapo as penance and as imitation of trajectories can also vary in determined zones, where
Jesus on His way to Golgotha. low density zones would result in long distance
During the Black Nazarene procession, the pedestrian trajectories, while high density zones
[7]
crowd densities reach extremes. Consequently, this would result in short pedestrian trajectories .
highly celebrated gathering attracts severe health Tracking pedestrians at high density crowd
and safety risks due to abnormal behavior of scenes is extremely challenging due to the small
pedestrians, such as panic stampedes. This has not number of pixels on the target, appearance
yet been systematically evaluated; ergo, risk vagueness resulting from packed condition, and
management efforts are constrained due to the lack critical inter-object occlusion. Ali and Shah (2008)
of information on the pedestrian characteristics developed an algorithm for tracking individuals in
during the event. high-density crowd situations. The algorithm uses a
For the crowd analysis, researchers scene structure-based force model in which an
developed tools that can be utilized to further risk individual, when moving in a scene, is subjected to
management efforts. Several variables were global and local forces that are functions of the
obtained to characterize the crowd dynamics, which layout of that scene and the behavior of other
include (1) time ratios, (2) relative pedestrian individuals in the same scene. The force models
density, (3) hustle rate, and (4) push formation. were captured by using the three floor fields, which
were inspired by the research in the field of
2. Literature Review evacuation dynamics. These floor fields are namely
Mass gatherings such as the Procession of Static Floor Field (SFF), Dynamic Floor Field (DFF)
the Black Nazarene exhibits different crowd and Boundary Floor Field (BFF). By combining all
characteristics to pedestrian traffic characteristics. the floor fields with the available information, it was
Karthika et al. (2018) discusses that in mass religious verified that using high level of knowledge in the
gatherings, crowd flows in multiple directions given scene using floor fields is helpful in improving the
[1]
one destination and does not stop even at large pedestrian tracking accuracy .
densities. In addition, a plausible cause for increased
crowd density is stagnant crowds. This effect causes 3. Methodology
unnecessary high crowd densities to occur The analytical survey method was used to
increasing the chances of a stampede occurring. record and assess the pedestrian dynamics during the
Due to the increasing occurrences of procession using video cameras.
stampedes during pedestrian crowding in mass
events such as pilgrimages, concerts, and parades, it 3.1 Equipment Specifications and Location
is timely to study the behavior of pedestrians to Selection
develop more refined models. This is in line with the The video cameras used were SJCAM
emergence of having more in-depth methods of SJ4000 WIFI units. The video quality is Full HD
analyzing pedestrian space-time dynamics to 1080P, generating 1920 by 1080-pixel frames at 30
improve necessary disaster risk and management frames per second.
measures during the said events. The study by Lian The road segments qualified for this study
et al. (2017) represented the movement of were situated along the procession route. An ocular
pedestrians in two ways. One of which is describing inspection of the procession route was done to
the movement of pedestrians as sequences described identify potential locations where the video cameras
in successive spatial lengths between two can be set up and where CCTV camera footages can
consecutive points through time. The second method be obtained.
describes the pedestrian movement as a binary point Video footage from 20 cameras were
process. Using video camera footages, the obtained for the study. It was necessary to observe
movements of pedestrians in a crowd were analyzed. multiple locations as the road width of the road
Through the distance-based representation, the sections that were being captured were considered.
Euclidean distance of a pedestrian between two

199