Machine Vision for X-Ray Image using Symbolic Mechanism

(1)

ISSN (Print) : 2278-8948, Volume-2, Issue-4, 2013

43

Machine Vision for X-Ray Image using Symbolic Mechanism

Akshay Agrawal¹& Anand Kr. Srivastava²

1Department of Electronic & Communication Engineering, ²Department of Computer Science & Engineering,

1&2JRE GROUP of Institutions, Greater Noida, India.

E-mail : [email protected]¹, [email protected]²

Abstract – Machine vision is the technology and methods used to provide imaging-based automatic inspection and analysis for such applications as automatic inspection, process control, and robot guidance in industry. The scope of MV is broad.MV is related to, though distinct from, vision. Machine vision methods are defined as both the process of defining and creating a MV solution, and as the technical process that occurs during the operation of the solution. In the following presented papers we hereby see the working of the MACHINE VISION through symbolic mechanism. While conventional (2D visible light) imaging is most commonly used in MV, alternatives include imaging various infrared bands, line scan imaging, 3D imaging of surfaces and X-ray imaging. Key divisions within MV 2D visible light imaging are monochromatic vs.

colour, resolution, and whether or not the imaging process is simultaneous over the entire image, making it suitable for moving processes. The most commonly used method for 3D imaging is scanning based triangulation. Other 3D methods used for machine vision are time of flight, grid based and stereoscopic.

Keywords – 3D, 2D, X-rays

I. INTRODUCTION

An Image contains a continuum of intensity values before it is quantized to obtain a digital image. The information is an image is in these gray values. To intercept an image, the variation in the intensity values shoul be analysed. In the early days when the machine vision was used then its availibility was really very much expensive and limited. The following limitations encounraged the designers of computer vision applications to focus their efforts and they hence created easy working with the binary system.

Even though the computers have become much more powerful, binary vision systems are still useful.

First of all, the algorithms for computing properties of binary images can be easily understood. A convenient wat to represent an object in gray level or color image is to use its mask . The mask of an object is the binary

picture in which the objects point is 1 and the other point is 0. After an object has been seperated from the background, its geometric and topological properties may be required in the decision making. Here the decision making refers to the process of copying the properties of the images and then comparing them.

II. REVIEW

THRESHOLDINGS: One of the most important problems in a vision system is to identify the sub images that represent objects. This operation, which is so natural and so easy for the people, is surprisingly difficult for computers. The partitioning of an image into various divided regions is termed as

“SEGMANTATION”. In other words segmentation can be defined as a method to partition an image, F [i,j], into sub images called regions, 𝑃1, … … . . 𝑃𝑘, such that each sub image is an object candidate,

Let us assume that a binary image B[i,j]is the same as a thresholdedgray image F_T[i,j]which is obtained using a threshold T for the original gray image F[i,j]. Thus,

B[i,j] = F_T[i,j] [1.1]

Where for a darker object on a lighter background F_T[i,j] = 1 if F[i,j] ≤ T

0 , otherwise [1.2]

Alit is known that the object intensity values are in a range [T1, T2], then we may use

F_T[i,j] = 1 if T1≤F[i,j] ≤ T2 0 , otherwise [1.3]

A general thresholding scheme in which the intensity levels for an object may come from several disjoint intervals may be represented as

(2)

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

44 F_T[i,j] = 1 if F[i,j]



Z

0 , otherwise [1.4]

Thresholding is the method to convert a grey scale image into a binary image so that the objects of interest can be separated from the background and can be used wherever you want to.

III. RUN LENGTH CODING

Another compact representation of a binary image is its run-length encoding. In this representation, numbers indicating the lengths of the runs of 1 pixels in the image are used to represent the image. This coding has been used for image transmission. Additionally, some properties, such as the area of objects. may be directly computed from their run-length codes.Two approaches are commonly used in run-length encoding. In the first, the start position and lengths of runs of is for each row are used. The other approach uses only lengths of runs, starting with the length of the 1 run. We will use the second convention and represent run lengths for each row of an image.

Thus, ri,k denotes the length of the kth run in the ith row of an image. The run-length codes for an image are shown in table

Start and length of 1 runs: (1,3) (7,2) (12,4)(17,2)(20,3) (5,13) (19,4)

(1,3) (17,6)

Length of 1 and 0 runs: 3, 3, 2, 3, 4, 1, 2,1, 3 0, 4, 13, 1, 4

3, 13, 6

Fig. 1 Run-length codes for a simple binary image.

The area of all objects can be obtained by summing the lengths of all 1 runs.

n-1 (mi-1)/2

A = Σ Σ ri,2k+1 [1.5]

i = 0 k = 0

Where m_i, is the number of runs in the ith row.

The horizontal projection can be easily computed from the run-length code, without generating the image (see Fig 2.8). The vertical and di agonal

projections can also be computed front the run-length code without generating the image by using clever code. Area and first and second mo ments can be calculated front the projections .

IV. BINARY ALGORITHM

Segmenting object pixels from background pixels is a difficult problem. We will not address this problem here.

Run Length Encode Horizontal projection (0,0)

(0,0) (8,6)

(4,5) (12,4) (3,1) (15,2) (2,1) (6,4) (15,2) (4,8) (14,2) (3,4) (10,6) (3,2) (12,2) (3,2) (6,2) (3,3) (2,5) (9,2) (1,1) (3,6) (1,1) (4,4) (11,3) (5,10)

(4,11) (3,1) (8,4) (4,2)

Fig. 2 : Horizontal Projection calculated from the run length code

V. SEQUENTIAL ALGORITHM

The sequential algorithm usually requires two Passes over the image. Since this algorithm works with only two rows of an image at a time, it can be used even when images are stored as a file and space limitations do not allow the full image to be brought into the memory.

In the algorithm there are three cases of interest when scanning an image left to right and top to bottom. The algorithm looks at only two of a pixel 4-neighbors: the ones above and then to the left of it. The algorithm will appear as follows:

1. Scan the image left to right, top to bottom.

2. If the pixel is 1, then

(a) If only one of its upper and left neighbors has a label, then copy the label.

(b) If both have the same label, then copy the label.

(c) If both have different labels, then copy the upper's label and enter the labels in the

(3)

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

45 equivalence table as equivalent labels.

(d) Otherwise assign a new label to this pixel and enter this label in the equivalence table.

3. If there are more pixels to consider, then go to step 2.

4. Find the lowest label for each equivalent set in the equivalence table.

5. Scan the picture. Replace each label by the lowest label in its equivalent set.

VI. MORPHOLOGICAL OPERATORS Mathematical morphology gets the following name from study of shape. This approach exploits the fact that in many machine vision applications, it is natural and easy to think in the term of shapes when designing algorithms. A morphological approach facilitates shape based, iconic or thinking. The fundamental unit of pictorial information in the morphological approach is the binary image.

The Intersection of any two binary images A and B written A ∩ B, is the binary image which is 1 at all pixels p which are 1 in both A and B. Thus,

A ∩B = {p|p € A and p € B}, [1.6]

The union of A and B, written A B , is the binary image which is 1 at all pixels p which are 1 in A or 1 in B (or 1 in both). Symbolically,

A B = {p|p € A and p € B}, [1.7]

Let Ω be a universal binary image (all 1) and A a binary image.

The complement of A is the binary image which interchanges the 1s and 0s in A. Thus,

A

= {p| p



Ω and p



A } [1.8]

The vector sum of two pixels p and q with indices [I,j] and [k,l] is the pixel p+q with indices [i+k, j+l]. The vector difference p-q is the pixel with indices [i-k,j-l].

VII. CONCLUSION

The use of x-rays as a light source for medical diagnosis has been around since the early 20th century.

The use of x-rays as a tool in non-destructive testing (for industrial use) has been around almost as long.

Photographic film has been the medium that converts the x-ray energy that it encounters into areas of light and dark depending on the amount of energy absorbed.

Referring to industrial use of x-rays only, many systems have been manufactured that replaced the photographic film with a combination of x-rays, fluorescent screen,

and a television camera. This type of system was mainly used as a non-destructive testing tool to inspect various products (such as tires) for internal flaws and always used an operator to make the decisions. In general, the images produced by such systems were poor in nature due to lack of contrast and noise. The improvements in digital image processing and complex algorithms have made it possible to combine machine vision and x-rays to address a whole new spectrum of applications that require automatic analysis for flaw inspection. The objective of this presentation will be to familiarize the audience with some of the techniques used to solve automatic real-time x-ray problems. References will be made to real applications in the aerospace, pharmaceutical, food, and automotive industries.

VIII. REFERENCES

[1] J. K. Aggarwal and N. Naudlutkumar. On the computation of motion from sequences of images: A review. Technical Report TR-88-2- 47. University of Texas. Austin. April 1988.

[2] Proc. SPIE 0850, Optics, Illumination, and Image Sensing for Machine Vision II, 45 (February 12, 1988); doi:10.1117/12.942859

[3] R. C. Bolles and M. A. Fischler. A RANSAC- based approach to model fitting and its application to finding cylinders in range data. In Int. Joint. Conf. Artificial Intelligence. pages 637 643. 1981.

[4] C. E. P. Box, W. G. Hunter. and J. S. Hunter.

Statistics for Ex-perimenters: An Introduction to Design, Data Analysis, and Model Building.

Wiley, New York. 1978.

[5] P...1. Hes' and R. C. Jain. Three-dimensional object recognition. Computing Surreys, 17(1):75 145. March 1985. [28) P. J. Fies1 and R. C.

Jain. Segmentation through variable-order sur- face fitting. IEEE Trans. Pattern Analysis and Machine Intelligence. 10(2):167-192, March 1988. (29) P. .1. Iles1 and N. D. McKay. A method for registration of 341 shapes. IEEE Trans. Pattern Analysis and Machine Intelligence, 14(2):239- 256. February 1992.

[6] S. T. Barnard and W. B. Thompson. Disparity analysis of images. IEEE Trans. Pattern Analysis and Machine Intelligence. 2(4):333-340, July 1980. [22) R. H. 13artels, J. C. Beatty. and B. A.

Barsky. An Introduction to Splines for Use in Computer Graphics and Geometric Modeling Nlor-gan Kaufmann. Las Altos. California, 1987.

[7] Bajesy. Active perception. Proc. IEEE, 76(8):996 1005. 1988.