YOLOv8 with Post-Processing for Small Object Detection Enhancement

Small-object detection in images, a core task in unstructured big-data analysis, remains challenging due to low resolution, background noise, and occlusion. Despite advancements in object detection models like You Only Look Once (YOLO) v8 and EfficientDet, small object detection still faces limitati...

Full description

Saved in:
Bibliographic Details
Main Authors: Jinkyu Ryu, Dongkurl Kwak, Seungmin Choi
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/13/7275
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849704362082828288
author Jinkyu Ryu
Dongkurl Kwak
Seungmin Choi
author_facet Jinkyu Ryu
Dongkurl Kwak
Seungmin Choi
author_sort Jinkyu Ryu
collection DOAJ
description Small-object detection in images, a core task in unstructured big-data analysis, remains challenging due to low resolution, background noise, and occlusion. Despite advancements in object detection models like You Only Look Once (YOLO) v8 and EfficientDet, small object detection still faces limitations. This study proposes an enhanced approach combining the content-aware reassembly of features (CARAFE) upsampling module and a confidence-based re-detection (CR) technique integrated with the YOLOv8n model to address these challenges. The CARAFE module is applied to the neck architecture of YOLOv8n to minimize information loss and enhance feature restoration by adaptively generating upsampling kernels based on the input feature map. Furthermore, the CR process involves cropping bounding boxes of small objects with low confidence scores from the original image and re-detecting them using the YOLOv8n-CARAFE model to improve detection performance. Experimental results demonstrate that the proposed approach significantly outperforms the baseline YOLOv8n model in detecting small objects. These findings highlight the effectiveness of combining advanced upsampling and post-processing techniques for improved small object detection. The proposed method holds promise for practical applications, including surveillance systems, autonomous driving, and medical image analysis.
format Article
id doaj-art-bd88edebb2bf4ef3ac731008d790698c
institution DOAJ
issn 2076-3417
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-bd88edebb2bf4ef3ac731008d790698c2025-08-20T03:16:46ZengMDPI AGApplied Sciences2076-34172025-06-011513727510.3390/app15137275YOLOv8 with Post-Processing for Small Object Detection EnhancementJinkyu Ryu0Dongkurl Kwak1Seungmin Choi2Department of Occupational Safety & Fire Protection, Woosuk University, Jincheon-gun 27841, Republic of KoreaGraduate School of Disaster Prevention, Kangwon National University, Samcheok-si 25913, Republic of KoreaGraduate School of Disaster Prevention, Kangwon National University, Samcheok-si 25913, Republic of KoreaSmall-object detection in images, a core task in unstructured big-data analysis, remains challenging due to low resolution, background noise, and occlusion. Despite advancements in object detection models like You Only Look Once (YOLO) v8 and EfficientDet, small object detection still faces limitations. This study proposes an enhanced approach combining the content-aware reassembly of features (CARAFE) upsampling module and a confidence-based re-detection (CR) technique integrated with the YOLOv8n model to address these challenges. The CARAFE module is applied to the neck architecture of YOLOv8n to minimize information loss and enhance feature restoration by adaptively generating upsampling kernels based on the input feature map. Furthermore, the CR process involves cropping bounding boxes of small objects with low confidence scores from the original image and re-detecting them using the YOLOv8n-CARAFE model to improve detection performance. Experimental results demonstrate that the proposed approach significantly outperforms the baseline YOLOv8n model in detecting small objects. These findings highlight the effectiveness of combining advanced upsampling and post-processing techniques for improved small object detection. The proposed method holds promise for practical applications, including surveillance systems, autonomous driving, and medical image analysis.https://www.mdpi.com/2076-3417/15/13/7275image processingcomputer visionobject detectionconvolutional neural networkYOLO
spellingShingle Jinkyu Ryu
Dongkurl Kwak
Seungmin Choi
YOLOv8 with Post-Processing for Small Object Detection Enhancement
Applied Sciences
image processing
computer vision
object detection
convolutional neural network
YOLO
title YOLOv8 with Post-Processing for Small Object Detection Enhancement
title_full YOLOv8 with Post-Processing for Small Object Detection Enhancement
title_fullStr YOLOv8 with Post-Processing for Small Object Detection Enhancement
title_full_unstemmed YOLOv8 with Post-Processing for Small Object Detection Enhancement
title_short YOLOv8 with Post-Processing for Small Object Detection Enhancement
title_sort yolov8 with post processing for small object detection enhancement
topic image processing
computer vision
object detection
convolutional neural network
YOLO
url https://www.mdpi.com/2076-3417/15/13/7275
work_keys_str_mv AT jinkyuryu yolov8withpostprocessingforsmallobjectdetectionenhancement
AT dongkurlkwak yolov8withpostprocessingforsmallobjectdetectionenhancement
AT seungminchoi yolov8withpostprocessingforsmallobjectdetectionenhancement