New Contrast Enhancement Techniques Based On Histogram Equalization Concept For Gray Scale Image
Even though the histogram equalization (HE) is well known for its simplicity and effectiveness in image contrast enhancement, nevertheless, it does suffer from excessive brightness change, intensity saturation and noise amplification problems. In general, the conventional HE based methods are divide...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://eprints.usm.my/43777/1/Lim%20Sheng%20Hoong24.pdf http://eprints.usm.my/43777/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Sains Malaysia |
| Language: | English |
| Summary: | Even though the histogram equalization (HE) is well known for its simplicity and effectiveness in image contrast enhancement, nevertheless, it does suffer from excessive brightness change, intensity saturation and noise amplification problems. In general, the conventional HE based methods are divided into two categories, namely brightness preservation (i.e., designed to preserve the brightness and improve the contrast of the natural images taken under normal lighting condition) and detail preservation (i.e., improve the visual and increase the brightness of the natural images taken under low lighting condition) methods. In this study, a brightness preservation method, namely Improved Quantized Plateau Limits Bi-Histogram Equalization (IQPLBHE) and a detail preservation method, namely Dynamic Range Bi-Histogram Equalization (DRBHE), are introduced. Basically, the proposed IQPLBHE method first separates the input histogram into two sub-histograms. Then, the plateau limits are calculated from the respective sub-histograms, which are used to modify those sub-histograms. Lastly, HE is then separately performed on the two sub-histograms. On the other hand, the proposed DRBHE method first separates the input histogram into two sub-histograms. Then, a new probability density function is created based on local information. Next, cumulative density function normalization is applied. Lastly, a new separating point is calculated before the combination of HE and DRHE is applied. Qualitative and quantitative analyses results show that both the proposed methods have good performance. Moreover, both the proposed methods have the advantages of being simple and tuning free, which are suitable to be applied in consumer electronic products. |
|---|