Contrast Enhancement by Nonlinear Diffusion Filtering

ABSTRACT:

To enhance the visual quality of an image that is degraded by uneven light, an effective method is to estimate the illumination component and compress it. Some previous methods have either defects of halo artifacts or contrast loss in the enhanced image due to incorrect estimation. In this paper, we discuss this problem and propose a novel method to estimate the illumination. The illumination is obtained by iteratively solving a nonlinear diffusion equation. During the diffusion process, surround suppression is embedded in the conductance function to specially enhance the diffusive strength in textural areas of the image. The proposed estimation method has the following two merits: 1) the boundary areas are preserved in the illumination, and thus halo artifacts are prevented and 2) the textural details are preserved in the reflectance to not suffer from illumination compression, which contributes to the contrast enhancement in the result. Experimental results show that the proposed algorithm achieves excellent performance in artifact removal and local contrast enhancement.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

• In existing filter to estimate the illumination, in which the weights were calculated by the frequency in the spatial domain of the relationship between the central pixel and its neighboring pixels. Another set of Retinex-like enhancement algorithms is based on the variational framework.

• These algorithms impose assumptions on both the reflectance and illumination, and they obtain both components by solving a quadratic optimization problem.

• Histogram equalization (HE) is one of the common methods used for improving contrast in digital images. However, this technique is not very well suited to be implemented in consumer electronics, such as television, because the method tends to introduce unnecessary visual deterioration such as the saturation effect. One of the solutions to overcome this weakness is by preserving the mean brightness of the input image inside the output image.

DISADVANTAGES OF EXISTING SYSTEM:

• Robustness of the proposed DHE against noises.

• Approach for contrast enhancement of low contrast images.

• Histogram splitting operation relieves the low frequency

PROPOSED SYSTEM:

• The flexibility and effectiveness of nonlinear diffusion exist in that its smoothing performance can be tuned by changing the equation formation and determining a different set of parameters. For example, it can smooth undesired information while respecting region boundaries and other structures in the image, as long as some crucial parameters are set appropriately.

• On the other hand, it tends to yield piecewise-constant-like images before it finally arrives at the globally constant solution.The above features are potential merits for estimating the expected illumination in this paper. In this section, the nonlinear diffusion is first introduced. Then, the texture suppression used to improve the smoothing properties for better handling of the expected illumination is described. Finally, the entire nonlinear diffusion method for illumination estimation as well as the acceleration strategy is presented in detail. The procedures of the proposed algorithm mainly involve illumination estimation, illumination-reflectance decomposition, illumination adjustment, and recombination of the two components. This type of structure addresses the illumination problem effectively and is widely adopted in variational Retinex algorithms.

• Note that our algorithm generally differs in two aspects: 1) the method for illumination estimation is via nonlinear diffusion; and 2) the mapping function for illumination adjustment is a logarithmic function. Considering that placing the enhancement on the three channels of the RGB color space potentially causes a “graying out” effect, the enhancement process should take place in the luminance component of the image to preserve the chrominance. In the proposed algorithm, a color space conversion from RGB to HSV is performed at the beginning and a reverse conversion at the end.

ADVANTAGES OF PROPOSED SYSTEM:

• This method solves the problem of the limited contrast enhancement that results from the low smoothing capability of a traditional nonlinear filter.

• The acceleration methods enable the algorithm to have high computational efficiency.

MODULES:

1. Image Acquisition
2. Illumination and Reflectance Decomposition
3. Mapping Function
4. Histogram Clipping

MODULES DESCRIPTION:

Image Acquisition:

• The first stage of any vision system is the image acquisition stage.

• The image has been obtained by Gallery.

Illumination and Reflectance Decomposition:

In the proposed algorithm, a color space conversion from RGB to HSV is performed at the beginning and a reverse conversion at the end.

In the decomposition process, the illumination component L (x, y)of the image luminance S(x, y) is first estimated using the nonlinear diffusion method, and then, the reflectance image R(x, y) is obtained by dividing S(x, y) by L(x, y).

Mapping Function:

Then, the illumination image L(x, y) is adjusted by a mapping function. Because the logarithmic function is strongly correlated to the nonlinear response of the human eye to the luminance, the adjustment is given as

L_a(x, y) = log₂ (L(x, y) + 1)

In practice, the log function has strong compression while preserving the brightness order, which makes the enhanced image look natural. This step is important for elevating the global brightness. The adjusted La (x, y) is then multiplied by the reflectance image R(x, y) again to obtain the adjusted luminance S _a (x, y):

S _a (x, y) = R(x, y) · La(x, y)

Histogram Clipping:

Next is the histogram clipping step, which cuts 0.5% of the pixels at the two ends of the histogram. This process can both remove the influence of a few pixels that have extreme luminance values and slightly enhance the image contrast. The last step is to rescale the dynamic range to [0 255] to obtain the enhanced luminance Se(x, y).

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

• System           : Pentium Dual Core.

• Hard Disk      : 120 GB.

• Monitor          : 15’’ LED

• Input Devices : Keyboard, Mouse

• Ram                : 1GB.

SOFTWARE REQUIREMENTS:

• Operating system : Windows 7.

• Coding Language : MATLAB

• Tool                       : MATLAB R2013A

REFERENCE:

Zhetong Liang, Weijian Liu, and Ruohe Yao, “Contrast Enhancement by Nonlinear Diffusion Filtering”, IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 25, NO. 2, FEBRUARY 2016.