Rozi, Atiqur (2026) CONTRAST ENHANCEMENT TECHNIQUES ON LUNG CT-SCAN IMAGES TO IMPROVE THE PERFORMANCE OF MOBILENETV2-BASED LUNG CANCER CLASSIFICATION. Undergraduate thesis, UPN Veteran Jawa Timur.
|
Text (Cover)
22081010010-Cover.pdf Download (1MB) | Preview |
|
|
Text (Bab 1)
22081010010-Bab 1.pdf Download (168kB) | Preview |
|
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Bab 2)
22081010010-Bab 2.pdf Restricted to Repository staff only until 25 May 2028. Download (569kB) |
|
|
Text (Bab 3)
22081010010-Bab 3.pdf Restricted to Repository staff only until 25 May 2028. Download (721kB) |
|
|
Text (Bab 4)
22081010010-Bab 4.pdf Restricted to Repository staff only until 25 May 2028. Download (2MB) |
|
|
Text (Bab 5)
22081010010-Bab 5.pdf Download (117kB) | Preview |
|
|
Text (Daftar Pustaka)
22081010010-Daftarpustaka.pdf Download (165kB) | Preview |
|
|
Text (Lampiran)
22081010010-Lampiran.pdf Restricted to Repository staff only Download (307kB) |
Abstract
Lung cancer remains the leading cause of cancer-related deaths globally, with delayed diagnosis frequently occurring due to asymptomatic conditions in the early stages. Although Computed Tomography (CT-Scan) is the gold standard for early detection, the resulting image quality often suffers from low contrast and visual noise, which hinders the accurate identification of nodules by intelligent systems. This study aims to comparatively analyze the effectiveness of four contrast enhancement techniques, namely Histogram Equalization (HE), Contrast Limited Adaptive Histogram Equalization (CLAHE), Gamma Correction, and Contrast Stretching, to overcome these issues. The performance of each method was evaluated based on its ability to increase accuracy and minimize the loss value in a Deep Learning-based lung cancer classification model using the MobileNetV2 architecture. In this study, the MobileNetV2 architecture was also evaluated using four hyperparameter configuration scenarios (frozen base and fine-tuning, as well as different learning rates). The results indicated that the application of contrast enhancement techniques can improve the performance of the MobileNetV2 architecture when compared to the baseline data, with Histogram Equalization (HE) emerging as the best method. The HE technique achieved the highest accuracy rate of 99% with a loss value of 1.20. In conclusion, the histogram equalization technique is recommended as the most optimal contrast enhancement technique. This is due to the technique's ability to sharply define the edges of cancer cell nodules, thereby improving the performance of the MobileNetV2 architecture in both the feature extraction and testing stages.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software Q Science > QA Mathematics > QA76.6 Computer Programming Q Science > QA Mathematics > QA76.87 Neural computers R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer) |
||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Informatics | ||||||||||||
| Depositing User: | Atiqur Rozi | ||||||||||||
| Date Deposited: | 26 May 2026 02:16 | ||||||||||||
| Last Modified: | 26 May 2026 02:16 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/52585 |
Actions (login required)
 |
View Item |