INTRODUCTION:

As the technology is growing people believe less in the manual work as it is tough task to maintain each and every record of a specific problem and as well as the security problems comes into mind when we talk about the manual data storage. Maintenance of manual data or paper based report is also a tough task, since it has to be kept safely from many external agents. Hence, here comes the need for a system which provides large security and less database maintenance.

Our paper basically focuses on this agenda only for the medical image (X-Ray image). Basically by developing this system the reports which are given from Doctors are embedded within the X-Ray image which helps in succeeding the idea of less database maintenance. For security purpose we can apply watershed method which is one of the cryptography method used for hiding the data [1]. Watershed is based on the encryption decryption method in which if some external agents try to access the data the watershed image breaks, break is nothing but a crack which helps the sender know that some external body or agent has tried to access the data.

LITERATURE SURVEY:

Generally complexity of method, resolution of image after data is hidden, the amount of data capable to be embedded define the quality of reversible data hiding technique. The quality of medical images differ from its diagnoses to diagnosis and so the technique results. Image quality define the criteria of how much data can be hidden. The correlation between the local block pixels in medical images is nearly zero because of its smoothness of surface. The Histogram shift method take advantage of this property of images to define the best threshold value. On the basis of the threshold the data embedding is categorized in two strategies i.e. negative and positive values. Data embedding strategy applies optimal threshold k to embed secret bits by a shift quantity nk (n ≥ 0). The basic idea of bit embedding is in specific range of threshold nk and –nk [2].

Another work talks about a novel scheme of distributed source coding. On the client side the information can be retrieved if and only if the receiver has the key with approximately same quality by applying image estimation algorithm. Image owner have to use the stream cipher technique for encrypting, than data compression is applied to remove the redundant bits from the encrypted image to make some space for data to hide. The taken out bits are encoded using Slepian-Wolf encoding using less density parity checking codes. If the receiver has both the embedding and encryption keys, he/she can extract the secret data and perfectly recover the original image using the distributed source decoding [3].

In order to preventing images from misuse and tempering the owner information (header) is embedded in the image, this method exploit the technique of Least Significant Bits in the area of interest and also safekeeping the changes made for further retrieval of original image. So the original pixels can be retrieved in very less distortion while the non area of interest pixels are retrieved in pseudo-random order. The testing is done in MRI and X-Ray images [4].

Encapsulating text in the colored image by changing pixel is a technique of Reversible Data Hiding, in which after encapsulation of data we can again retrieve it. It works on very high resolution image for making the data hiding reversible, Also after the data is embedded, image can be recovered later. The old Lee and Pixel Pair Mapping (PPM) method of DE gives various places for data hiding by clustering the two columns of pixel values, but the updated one can cluster any number of columns pixel values in onwards orientation and rearwards orientation. The difference duo match in onward and rearwards orientation exploits the prediction method using random threshold, and the predicted results of pixel duo provides the security [5]. Authors in [6,7] performed blind data hiding in X-ray, MRI images.

In the segmentation scheme we have followed the traditional way of segmentation because the images are prone to be distorted due to the effect of noise and surroundings. This algorithm works on the method of edge detection, image preprocessing and selecting the co-ordinates [8].

In the wavelet transformation the segmentation is done by the contrast of the image. The section which lies in the contrast is segmented from the area that does not lie in the contrasted area. On the basis of the texture of the image we forms the matrix that the areas of the segmentation on which we have to perform the embedding.

In this method of dividing ROI and RONI the segmentation is done by dividing the image into two regions that is region of interest and region of non interest. For the segmentation purpose we have chosen the X-Ray images in which we have divided the image on the basis of the gray scale.

This region of interest can be chosen by defining the co-ordinates in the image. All the area that is covered by the co-ordinates is called as the region of interest and the area that belongs to the outside if the co-ordinates is called as the region of non interest.

It is found in the literature [9-11] that, Cloud computing is gaining popularity now a days in providing its infrastructure, security, storage as services to the end user. In this paper we develop a client server based architecture model for maintaining X-Ray images with patient information embedded in it. The method makes the record keeper job easy and simple in any hospital or clinical laboratory.

METHODOLOGY:

The methodology includes the following steps in our system. Fig. 1. shows the steps involved in the design of the proposed system.

Preprocessing:

In this we pre-process the image and improve the quality of the image. For the preprocessing we have used image processing and edge detection. In the preprocessing step we have done the image sharpening, edge detection and boundary filling then we have applied the segmentation algorithm to find the region of interest. For the text embedding we have used the LSB.

Segmentation:

In this we segment the image into two regions that is ROI and RONI. Region of interest includes the area in which we embed the text. Fig. 2 shows the steps involved in the segmentation process.

Text-Embedding:

In the segmented image we embed the text in the region of interest. The goal is to embed the header of the image and other data in the image itself by following the inseparability and tamper-proofing capabilities. To perform this step, we applied the data hiding technique proposed in [7] using LSBs hides information into the ROI and preserves lost information in the RONI for reversibility purposes.

ROI pixels are accessed consecutively or in accordance to a Just Noticeable Distortion (JND) model, while RONI pixels are accessed in a consecutive or pseudo-random order. In the experiments with MRI and X-Ray images [7]. Fig. 3. shows the steps involved in the text embedding process.

Fig. 1. Architecture Diagram

Fig. 1. Architecture diagram for Segmentation Process.

Fig. 2. Architecture Diagram for Text Embedding Process.

Fig. 3. Client Portal for Uploading X-Ray Image.

RESULTS AND ANALYSIS:

Fig. 4. shows the client and GUI interface for the proposed model. Fig. 5. shows the GUI design of the proposed system for embedding text in X-Ray image.

In the proposed system we are going to compare three of the segmentation algorithm they are fuzzy-c mean, watershed and histogram.

Table 1. shows the analysis and the comparison of different segmentation methodologies, with the output of each algorithms, for separating RIO from RONI. It includes Histogram, Fuzzy C-Mean, Watershed algorithm, in which we found best results for watershed algorithm.

Fig. 4. GUI Design for Embedding Text in X-Ray Image.

CONCLUSION:

We have developed a system which basically helps in reducing the database maintenance and also help in the security purpose by embedding text into the X-Ray image. In future we can try to apply watershed algorithm over the provided image so as to provide secure transfer of file over the internet.

REFERENCES:

1. Zhang, X., Jia, F., Luo, S., Liu, G., and Hu, Q. (2014). A marker-based watershed method for X-Ray image segmentation. Computer methods and programs in biomedicine, 113(3), 894-903.

2. Huang, L. C., Tseng, L. Y., and Hwang, M. S. (2013). A reversible data hiding method by histogram shifting in high quality medical images. Journal of Systems and Software, 86(3), 716-727.

3. Qian, Z., and Zhang, X. (2015). Reversible Data Hiding in Encrypted Image with Distributed Source Encoding.

4. Fylakis, A., Keskinarkaus, A., Kiviniemi, V., and Seppanen, T. (2015, March). Reversible blind data hiding for verifying integrity and authenticating MRI and X-Ray images. In Medical Information and Communication Technology (ISMICT), 2015 9th International Symposium on (pp. 185-189). IEEE.

5. inoth, K. C., Natarajan, V., and Santhosh, M. S. (2014, April). Difference expansion based Reversible Data Hiding for medical images. InCommunications and Signal Processing (ICCSP), 2014 International Conference on (pp. 720-723). IEEE.

6. Schlüter, S., Sheppard, A., Brown, K., and Wildenschild, D. (2014). Image processing of multiphase images obtained via X‐ray microtomography: a review. Water Resources Research, 50(4), 3615-3639.

7. Angelos Fylakis, Anja Keskinarkaus, Vesa Kiviniemi, Tapio Seppanen. "Reversible blind data hiding for verifying integrity and authenticating MRI and X-Ray images", 2015 9th International Symposium on Medical Information and Communication Technology (ISMICT), 2015.

8. Ram, S., and Rodriguez, J. J. (2013, May). Random walker watersheds: A new image segmentation approach. In Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on (pp. 1473-1477). IEEE.

9. Zeng, W., Zhao, Y., Ou, K., and Song, W. (2009, November). Research on cloud storage architecture and key technologies. In Proceedings of the 2nd International Conference on Interaction Sciences: Information Technology, Culture and Human (pp. 1044-1048). ACM.

10. Kondo, D., Javadi, B., Malecot, P., Cappello, F., and Anderson, D. P. (2009, May). Cost-benefit analysis of cloud computing versus desktop grids. InParallel and Distributed Processing, 2009. IPDPS 2009. IEEE International Symposium on (pp. 1-12). IEEE.

11. Rimal, B. P., Choi, E., and Lumb, I. (2009, August). A taxonomy and survey of cloud computing systems. In 2009 Fifth International Joint Conference on INC, IMS and IDC (pp. 44-51). IEEE.

Received on 13.07.2016 Modified on 22.07.2016

Research J. Pharm. and Tech 2016; 9(9):1337-1340.

DOI: 10.5958/0974-360X.2016.00255.9

S. NO.	SEGMENTATION ALGORITHM	RESULT
1.	Histogram: The result produced by the histogram method doesn’t identify the region of interest properly.
2.	Fuzzy-c mean: The output given by this method is not able to separate the blurred non-region of interest from region of interest.
3.	Watershed : The watershed algorithm gives the best output since it properly gives the region of interest as well as the non- region of interest.