EVANDRA, HENRY and Darmawahyuni, Annisa (2024) DETEKSI OBJEK LIGHTWEIGHT LOW IMAGE MENGGUNAKAN ARSTITEKTUR YOLOv5. Undergraduate thesis, Sriwijaya University.
![[thumbnail of RAMA_55201_09021282126074.pdf]](http://repository.unsri.ac.id/style/images/fileicons/text.png) |
Text
RAMA_55201_09021282126074.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) |
|
Text
RAMA_55201_09021282126074_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) |
|
Text
RAMA_55201_09021282126074_0030069005_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (983kB) |
|
Text
RAMA_55201_09021282126074_0030069005_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (658kB) |
|
Text
RAMA_55201_09021282126074_0030069005_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (361kB) |
|
Text
RAMA_55201_09021282126074_0030069005_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (508kB) |
|
Text
RAMA_55201_09021282126074_0030069005_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) |
|
Text
RAMA_55201_09021282126074_0030069005_06.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (180kB) |
|
Text
RAMA_55201_09021282126074_0030069005_07_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (179kB) |
|
Text
RAMA_55201_09021282126074_0030069005_08_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (254kB) |
Abstract
Object detection is the process of identifying and localizing a specific object in an image or video that aims to recognize the presence and position of the object specifically. This technology has various applications in daily life, such as face detection to open smartphones. However, a major challenge in object detection is the camera's sensitivity to light intensity. In low lighting conditions, images may suffer from low contrast, resulting in blurry and distorted images. This condition can affect the performance of computer vision-based object detection systems. This research proposes the YOLOv5 model in low-light object detection systems. From the analysis, the YOLOv5x version of the model shows the best performance but has a larger size and complexity. Therefore, the YOLOv5m model with a confidence threshold of 0.3 is selected as a more efficient solution. This model offers a balance between precision, recall, and mAP, and has a moderate size compared to the other versions, making it suitable for object detection applications under various lighting conditions.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | Citra Gelap, Deteksi Objek, YOLOv5 |
| Subjects: | T Technology > T Technology (General) > T1-995 Technology (General) |
| Divisions: | 09-Faculty of Computer Science > 55201-Informatics (S1) |
| Depositing User: | Henry Evandra |
| Date Deposited: | 30 Dec 2024 07:32 |
| Last Modified: | 30 Dec 2024 07:32 |
| URI: | http://repository.unsri.ac.id/id/eprint/161985 |
Actions (login required)
 |
View Item |