Analyzing the Effect of JPEG Compression on Local Variance of Image Intensity

ABSTRACT:

The local variance of image intensity is a typical measure of image smoothness. It has been extensively used, for example, to measure the visual saliency or to adjust the filtering strength in image processing and analysis. However, to our knowledge, no analytical work has been reported about the effect of JPEG compression on image local variance. In this paper, a theoretical analysis on the variation of local variance caused by JPEG compression is presented. First, the expectation of intensity variance of 8_8 non-overlapping blocks in a JPEG image is derived. The expectation is determined by the Laplacian parameters of the DCT coefficient distributions of the original image and the quantization step sizes used in the JPEG compression. Second, some interesting properties that describe the behavior of the local variance under different degrees of JPEG compression are discussed. Last, both simulation and experiments are performed to verify our derivation and discussion. The theoretical analysis presented in this paper provides some new insights into the behavior of local variance under JPEG compression. Moreover, it has the potential to be used in some areas of image processing and analysis, such as image enhancement, image quality assessment and image filtering.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

• A few works have been reported on the analysis of local variance of image intensity. In order to study the block-based discrete cosine transform (block-DCT for short) coefficients of real images, Lam and Goodman analyzed the distribution of local variance. In their work, the local variance was assumed to follow an exponential or half-normal distribution.

• Troncoso-Pastoriza and Perez-Gonzalez proposed a skewed log-stable model to fit the distribution of local variance, and further presented an efficient method to estimate the parameters of the log-stable model. Both of the works mainly concern the distribution of the local variance of uncompressed images. We also notice that in the filed of image quality assessment, a great deal of effort has been devoted to assess the perceptual quality of images by measuring the distortion caused by JPEG compression

DISADVANTAGES OF EXISTING SYSTEM:

No analytical work has been done to study the behavior of local variance under JPEG compression.

PROPOSED SYSTEM:

• In this paper, we focus on analyzing the effect of JPEG compression on local variance. Under the assumption that the alternating-current (AC) coefficients of the block-DCT follow a Laplacian distribution, we first derive a theoretical expression for the expectation of local variance.

• Based on the derivation, we then analyze the variation of local variance caused by JPEG compression.

• Finally, simulation and experiments are conducted to verify our derivation and analysis.

ADVANTAGES OF PROPOSED SYSTEM:

• The expectation of local variance is determined by the quantization step sizes of JPEG compression and the distribution parameters of AC coefficients of the original uncompressed image.

• The variation of local variance caused by JPEG compression has been quantified by the ratio of expectation of local variances between before and after JPEG compression.

MODULES:

1. JPEG Compression

2. Formulation of Local Variance

3. Derivation of Expectation of Local Variance

4. Analysis of Variation of Local Variance

MODULE DESCRIPTION:

JPEG Compression:

• JPEG compression consists of three basic steps: block-DCT, coefficient quantization, and entropy coding. The degree of JPEG compression is partially affected by the quantization table (QT, i.e., a matrix of quantization step sizes). According to the JPEG standard, the quantization table can be configured by an individual user.

• Based on JPEG Compression, in our work compressed image is generated.

Formulation of Local Variance:

In this stage, formulate the local variance for original and compressed image. First take original image for formulate the local variance. In original image apply DCT for 8*8 blocks. In this process, we are get DC and AC coefficients. DC coefficient is zero remaining has 63 AC coefficients for find local variance. Then find local mean for AC coefficients. Local mean means sum of AC coefficients. Finally get local variance by sum of square of AC coefficients.

After formulate the local variance of original image, take compressed image for formulate the local variance. In compressed image apply DCT for 8*8 blocks. In this process, also we are get DC and AC coefficients. DC coefficient is zero remaining has 63 AC coefficients for find local variance. AC coefficients quantized by quantization step size after extract AC coefficients. Finally get local variance by sum of square of quantized AC coefficients.

Derivation of Expectation of Local Variance:

In this stage, derive the expectation of local variance for original and compressed image. Calculate the expectation of the local variance of a JPEG image under the assumption of Laplacian distribution.

Analysis of Variation of Local Variance:

In this stage, variation can be quantified by the ratio of the expectation values of the local variances before and after JPEG compression. Finally we are get local variance is high for highly compressed image [QF is above 90].

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:

Jianquan Yang, Guopu Zhu, Senior Member, IEEE, and Yun-Qing Shi, Fellow, IEEE, “Analyzing the Effect of JPEG Compression on Local Variance of Image Intensity”, IEEE Transactions on Image Processing 2016.