Detecting vehicles using the modified YOLOv۴ algorithm withhigh accuracy in deep learning

The abstract is to be in fully-justified italicized text as it is here, belowthe author information. Use the word “Abstract” as the title, in ۱۲-pointTimes New Roman, boldface type, centered relative to the column,initially capitalized. The abstract is to be in ۱۲-point, single-spacedtype. The abstract should give a concise and informative description ofthe paper, between ۱۵۰ to ۳۰۰ words. All manuscripts must be inEnglish. (۱۲ pt) Vehicle detection from images is an importantapplication in various fields, including military, security, urbansurveillance, transportation, and safety. Deep learning algorithms havebecome a popular choice for this task due to their high accuracy andefficiency. In this paper, two modified algorithms of YOLOv۴ arepresented for vehicle detection from images. These algorithms, throughseveral modifications, significantly increase the detection accuracy.YOLOv۴ is a fast and efficient object detection algorithm based onconvolutional neural networks. This algorithm divides the image intosmall regions and assigns a label to each region, which can include thetype, position, and size of the object. These modified algorithms wereevaluated on a large dataset of images. The results show that thesealgorithms significantly increase the accuracy and speed of detection.The results of this paper demonstrate that deep learning algorithms can significantly improve the accuracy of vehicle detection from images.These modified algorithms can be used for various applications indifferent fields, including military security, urban surveillance,transportation, and safety

کلیدواژه ها:

Vehicle detection ، deep learning algorithm ، YOLOv۴ ، improved algorithm