TY - JOUR
T1 - Air Vehicle Classification System and Speed Alert for the Prevention of Accidents on Flat and Curved Roads
AU - Pastrana, Estefany Lorenzo
AU - Curasma, Joshy Nichelson Castillo
AU - Tovar, Esthefany Araceli Curo
AU - Joija, Jimmy Nick Stevens Garcia
N1 - Publisher Copyright:
© 2024 by authors, all rights reserved.
PY - 2024/9/1
Y1 - 2024/9/1
N2 - According to the MTC, there were 65,539 M20 speeding offenses in September 2022, increasing the number of traffic accidents by 3312 people killed. This work develops an aerial classification system to alert speeding on flat and curved roads. To obtain the data in the Breña bridge and the pedestrian bridge of the National University of Central Peru (UNCP) in Huancayo, it was subdivided into area 1 and area 2, where the Mavic Air drone was used to implement the vehicle classification system YOLOv7. It was trained with the COCO base and then added to the Python libraries where the programming was performed using the dynamic behavior of a particle at constant speed to track the vehicle to calculate the centroid of the rectangle and the speed that travels on the road. Finally, it was obtained from the 10 samples considered for each scenario in the detection of speeding that 60% of the cars travelling on the Breña bridge exceeded the speed limit, while the vehicles evaluated in area 1 exceeded the speed limit of 30km/h in the school zone 100%, while in area 2 80% of cars exceeded the permitted limit, indicating that the speeding alert is required in the evaluated sites to adequately alert the driver when entering this stretch of road.
AB - According to the MTC, there were 65,539 M20 speeding offenses in September 2022, increasing the number of traffic accidents by 3312 people killed. This work develops an aerial classification system to alert speeding on flat and curved roads. To obtain the data in the Breña bridge and the pedestrian bridge of the National University of Central Peru (UNCP) in Huancayo, it was subdivided into area 1 and area 2, where the Mavic Air drone was used to implement the vehicle classification system YOLOv7. It was trained with the COCO base and then added to the Python libraries where the programming was performed using the dynamic behavior of a particle at constant speed to track the vehicle to calculate the centroid of the rectangle and the speed that travels on the road. Finally, it was obtained from the 10 samples considered for each scenario in the detection of speeding that 60% of the cars travelling on the Breña bridge exceeded the speed limit, while the vehicles evaluated in area 1 exceeded the speed limit of 30km/h in the school zone 100%, while in area 2 80% of cars exceeded the permitted limit, indicating that the speeding alert is required in the evaluated sites to adequately alert the driver when entering this stretch of road.
KW - Centrifugal Force
KW - Drone
KW - Vehicle Classification
KW - Vehicle Detection
KW - Vehicle Speed
UR - http://www.scopus.com/inward/record.url?scp=85201569867&partnerID=8YFLogxK
U2 - 10.13189/cea.2024.120501
DO - 10.13189/cea.2024.120501
M3 - Original Article
AN - SCOPUS:85201569867
SN - 2332-1091
VL - 12
SP - 3139
EP - 3150
JO - Civil Engineering and Architecture
JF - Civil Engineering and Architecture
IS - 5
ER -