Description
Method and system for increasing image contrast
The present invention relates to a method and system for increasing the contrast of an image of a surface.
Image display systems emit light to a surface of interest to create an image of the surface. The light is formed of an array of picture elements or pixels having a particular brightness or intensity value, which together form a viewable image when captured by a camera. The contrast of an image is regarded-as the difference in brightness between the light and dark areas, i.e. the brightness distribution of the image.
Depending of the lighting of the surface dark areas might appear as almost black in the image, although the area of the surface has a structure which might be of interest, e.g; for a surface inspection. In other words the contrast of the part of the image with this area is very low. The same can happen , with bright areas which appear almost white in the image. This is a common problem, since every camera has a limited resolution of brightness levels. The visual appearance of the image can often be improved by adjusting the brightness of the pixels, which is called as contrast enhancement.
Many techniques for contrast enhancement and modification have been proposed in the prior art. Gamma correction and histogram equalization are examples of methods of correcting low contrast. However these methods are limited in their ability of restoring a high contrast to all areas of the image.

The object of the invention is to decrease areas of low contrast within an image.
These problems are solved by a method according to claim 1 and a system according to claim 10.
By locating the projector pixels corresponding to a low contrast portion of the image of a surface and by adjusting the brightness of these projector pixels accordingly, the contrast of the low contrast portion can be increased locally. This local adjustment of the brightness avoids an overall adjustment of the brightness, which could lead to other areas being of low contrast. Thus, this invention allows to overall decrease areas of low contrast in an image. The local contrast enhancement helps to obtain high quality images of the surface, for example to inspect for surface defects with increased precision.
According to a preferred embodiment, a map is used to locate the projector pixels corresponding to the identified portion of the image. This map provides exact location of the pixels with no additional computation, thereby making it easy to locate the projector pixels.
According to a preferred embodiment, the map is obtained by projecting the set of projector pixels in a predefined geometric pattern to the surface, obtaining a corresponding image of the surface and identifying the location of the projector pixels of the predefined geometric pattern in the image. The predefined geometric pattern helps to easily identify the corresponding pixels in the image with higher precision.

According to a preferred embodiment, the predefined geometric pattern is a striped pattern. This pattern can easily be recognized in the image and allows to easily identify the location of the projector pixels. This way the map can easily be generated.
According to another preferred embodiment, the predefined geometric pattern is a checker board pattern. This pattern makes it easy to distinguish the projector pixels and its location in the image.
According to a preferred embodiment, the mapping is performed for the identified portion of a predefined size. This allows increasing the local contrast of the dark areas without any artifacts which was there if the size is too small. Also, it helps to fix a size that balances the locality of the contrast enhancement.
According to a preferred embodiment, adjusting the-brightness includes increasing the brightness of the identified portion if the brightest camera pixels in the identified portion are darker than an identified threshold. This enables secure identification of the dark areas of the image and to adjust the brightness to increase its contrast using a specific threshold.
According to yet another preferred embodiment, adjusting the brightness of the identified portion includes decreasing the brightness of the identified portion if the darkest camera pixels in the identified portion are brighter than the identified threshold. This allows to identify the white areas of the image and to increase its contrast using a specific threshold.

According to a preferred embodiment, the method is used to detect surface defects of the surface. The contrast of the image of the surface is enhanced locally by adjusting the brightness of the pixels to fit to a particular area based on the comparison of the image pixel brightness to a detected brightness level. This high contrast image makes it easier to detect the presence of any surface defects with increased precision. The local adjustment of contrast helps to avoid artifacts on the image.
According to another preferred embodiment, the projector includes at least one of a micromirror device and a liquid crystal display to project the set of pixels to the surface.
According to another preferred embodiment, the camera is at least one of a CCD camera and a CMOS camera. The CCD camera is capable of capturing the pixels even of pixel sizes down to a few micrometres in real time to provide a clear image. CMOS camera offers more-pixel integration, lower power dissipation, and the possibility of smaller system size.
The present invention is further described hereinafter with reference to illustrated embodiments shown in the accompanying drawings, in which:
FIG 1 schematically shows a system for increasing
contrast of an image of a surface, according to an embodiment herein,
FIG 2 shows a three-dimensional view of the arrangement of the system of FIG.l,

FIG 3A-3D illustrates schematic diagrams depicting examples
of predefined geometric pattern of pixels projected to the surface,
FIG 4A-4B shows an exemplary illustration of measuring
characteristics of the surface according to an embodiment of the invention,
FIG 5A-5C illustrates a method of mapping the projector pixels with the image pixels according to an embodiment of the invention,
FIG 6A-6B shows an exemplary illustration of increasing the contrast of low contrast object in the surface,
FIG 7A-7D shows another exemplary illustration of increasing the contrast of an object according to an embodiment herein, and
FIG 8 illustrates a flow diagram depicting a method of
increasing the contrast of an image according to an embodiment of the invention.
Various embodiments are described with reference to the drawings, wherein like reference numerals are used to refer to like elements throughout. In the following description, for purpose of explanation, numerous specific details are set forth in order to provide a thorough understanding of one or more embodiments. It may be evident that such embodiments may be practiced without these specific details.
FIG 1 schematically shows a system 10 for increasing contrast of an image 15 of a surface 13, according to an embodiment herein. As illustrated in figure, the system 10 comprises of

a projector 11, an imaging camera 14 and a processing means 17. The projector 11 allows projection of a set of projector pixels 12 corresponding to various lighting effects such as images, patterns, etc onto a surface 13. The surface 13 on to which the set of pixels 12 are projected is generally three-dimensional in nature and includes a plurality of objects of different contrasts and which are arranged at various depths with respect to the camera 14. The projector 11 includes a light source which functions as a spatial modulator for the projector pixels 12. Preferably, the light source is a micromirror device 19 comprising a plurality of mirrors 20 arranged in a matrix. Here each mirror 20 represents'one or more pixels 12 in the projector 11 which can be driven sequentially to modulate the distribution of pixels 12 to the surface 13.
The number of mirrors 20 corresponds to the resolution of the projector 12. The projector pixels 12 can be repositioned by ■■ rapidly toggling the mirrors 20 on- and off -to ■produce grayscales controlled by the ratio of the. on time and off time. The projector 10 can also be an LCD panel to project pixels 13 of different sizes to the surface 14. For instance, 800x600, 1024x768, 1280x7.20, 1920x1080 matrices are some common pixel sizes.
The set of projector pixels 12 transmitted from the projector 11 incidents and reflects off at the surface 13 and reaches the camera 14. The camera 14 captures an image 15 of the surface 13. As the surface 13 herein includes different objects at different depths and contrasts, the projector pixels 12 incident at different points of the surface 13 reflects to reach different pixels on the camera 14. Here, the camera 14 may be charge coupled device (CCD) type camera, a CMOS video camera or the like.

The image 15 captured by the camera 14 is then examined to identify a portion 18 of the surface 13 with a low contrast in the image 15. The processing means 17 associated with camera 14 further locates the projector pixels 12 which are projected to a part of the surface 14 and are captured in the identified portion 18 of the image 15. The processing means 17 then creates a three-dimensional model 21 of the surface 13 by a method of triangulation using the projector 11, the pattern on the surface 13 and the camera 14 as points of reference. Further, a map to locate the projector pixels 12 corresponding to the identified portion 18 of the image 15 is framed based on the three-dimensional model 21. Here, the map and the three dimensional model 21 corresponds to each other. The map is obtained by projecting the set of projector pixels 12 in a predefined geometric pattern 22 to the surface 13, obtaining a corresponding image 15 of the surface 13 and further locating the position of projector pixels 12 in the image 15
The camera 14 captures projections of known patterns, and correspondences are found between image pixels 16 and projector pixels 12. If the camera 14 is calibrated, and the display surface 13 geometry is known, then the projector 11 calibration parameters can be determined. From the information from the camera 14, a three-dimensional model 21 corresponding to a map of the surface 13 can be obtained using triangulation method.
The corresponding points between the pixels 12 in the projector 11 and the image 15 captured by the camera 14 can thus be established using an interpolated map created by codification of the three-dimensional model 21, thereby

obtaining the brightness pattern for projecting the pixels 12 to the surface 13.
The processing means 17 further provides instructions to adjust the brightness of the located projector pixels 12 so as to increase the contrast of the identified portion 18 of the image 15. The brightness of pixels 16 at the identified portion 18 of the image 15 is noted.and is then compared with the brightness required for increasing the contrast. Here, adjusting the brightness refers to increasing the brightness of the identified portion 18 if the brightest image pixels 16 in the identified portion are darker than an identified threshold. The adjusting of brightness of the identified portion may also be performed by decreasing the brightness of the identified portion 18 if the darkest image pixels 16 in the identified portion 18 are brighter than the identified threshold.
-For adjusting the brightness of-the projector pixels 12, the differences in brightness of the identified portion and brightness required for increasing the contrast is added to the gray levels in the projector pixels 12 to subsequently increase or decrease the brightness of the set of projector pixels 12 projected to the surface 13 to equalize the brightness to the desired brightness. Thus a brightness corrected brightness pattern is obtained for increasing the contrast of the identified portion 18.
The corrected brightness pattern is then impinged on the surface 13 and the high contrast image 15 is captured using the camera 14. Further, the image 15 is processed by the processing means 17, for instance, a digital signal processor to evaluate for surface defects. Here the contrast of the pixels corresponding to the low contrast portions of the

images is suitable adapted to increase the brightness such that it does not create any artifacts, especially in regions which are too dark or too light.
One embodiment of the system 10 includes a projector 11 provided with an array of micro mirrors 20, which are arranged in multiple geometric segments so that the plurality of mirrors 20 of one segment in the array can be toggled on and off to project pixels 12 at various brightness to optimize the image 15 of the surface 13 being captured by the camera 14 for inspection.
FIG. 2 illustrates a three-dimensional view of the system of FIG.l. The projector 11 herein emits an array of pixels 12 to the surface 13. The surface 13 herein in generally three-dimension and includes a plurality of objects 29 of various sizes and shapes. The arrows herein indicate the direction of travel of pixels 12.
The pixels 12 incident on the surface 13 is reflected off from the surface 13 and a camera 14 is positioned to receive the reflected pixels 12. The camera 14 captures the pixels 14 and creates an image 15 of the surface 13 corresponding to the distribution of the pixels on the surface of the objects 29.
The image 15 is then analyzed by the processing means 17 using an algorithm to provide for the three-dimensional geometry of the surface 13. The algorithm takes into account the contrast of the different objects 29 in the image 15.
FIG 3A-3D illustrates a schematic diagram depicting examples of predefined geometric pattern 22 of projector pixels 13 projected to the surface 13. The predefined geometric pattern

22 of the pixels 13 can be a striped pattern 23 as shown in 3A. Concurrently, the FIGS 3B, 3C, and 3D show projector pixels 12 arranged in a checker board pattern 24, triangular pattern 25 and diagonal pattern 26.
The process of projecting known pattern of pixels 12 to a surface is termed as structured lighting. One example of such a system which utilizes structured lighting is known from US20050157920. The objects in a scene are individually identified and the structured light.pattern is aligned for each object.A controller associated with the system dynamically adjusts parameters for structured illumination according to the actual position of the object being inspected.
The pattern 22 that the pixels 12 deform when striking on the surface 13 allow to calculate the depth and other characteristics of the surface 13. These patterns 23-26 are generated by the light source within the projector 11, typically an LCD or a micromirror device 19. The predefined geometric pattern 22 generally used is a striped pattern 23 as it have high resolution and good accuracy.
The geometric arrangement of the projector pixels 12 varies depending on the usage. In monitors, such as LCDS or CRTs that typically display edges or rectangles, the pixels 12 are arranged in vertical stripes. The displays with motion pictures may have a triangular pattern 25 or diagonal pattern 26 so as to have a better perception of the image 12 variation by the user.
A digital camera 14 and processing means 17 can be used to calibrate the projector 11 with the assistance of predefined geometric patterns 22. To account for distortion effects and

to equalize brightness, it is desirable to know the mapping of the points within each projector pixel 12. A small set of pixels 12 can be projected from each projector 11 element if a predefined pattern 22 is used and can be recognized in the camera image 16.
FIG 4A-4B shows an exemplary illustration of measuring characteristics of the surface 13 according to an embodiment herein. The example shows the patterns of pixels 12 projected to and reflected off from the surface 12. The projector pixels 12 are impinged to the surface in a binary striped pattern 23 in a horizontal manner as shown in FIG. 4A. The micromirrors 20 in the projector 11 are arranged in a binary grating pattern to project a set of pixels 12 in a horizontal striped pattern 23 to the surface 13 to be inspected. The pixels 12 reflecting from the surface 13 to the camera 14 shows a deformation in the pattern as shown in FIG 4B. The objects present in the surface 13 impose the deformation to the pixels 12. The pixels 12 reflected to the camera 14 exhibits deformations dependent on the depth of the objects in the surface 13.
Projecting the pixels 12 in a striped pattern 23 onto a three dimensionally shaped surface 13 produces a line of illumination that appears distorted from other perspectives than that of the projector 11, and can be used for an exact geometric reconstruction of the surface 13 shape. The results show very high accuracy, capturing even smallest surface 13 details. The acquired maps can be further processed using standard techniques to produce a complete 3D model 21 of the surface 13.
A structured pattern of pixels 12 can be provided by any programmable lightning means which can be dynamically

controlled. For example, a selectively transmissive device such as an array of pixels selectively allowing transmission of light beam may be used as a projector. FIG 5A-5C illustrates a method of mapping the projector pixels 12 with the image pixels 16 according to an embodiment of the invention. FIG 5A shows an image 15 of the surface 13 impinged with a set of projector pixels 12 to illuminate the surface 13 to be inspected. The surface. 13 includes multiple objects for instance, a resistor 27 and a thermocol pyramid 28 having various depths and; contrasts. The image 15 clearly indicates that some parts of the. surface 13 appear dark and some other parts light due to different surface areas prevalent at different distances.
The image 15 of the surface 13 obtained when impinged with a set of projector pixels 12 in a binary striped pattern 23 is as shown in FIG. 5B. The resistor 27 and the pyramid 28 on the surface 13 produce deformation on the incident projector pixels- 42>' thereby providing information on correspondences and other characteristics of the surface 13. The image 16 shows several depth cues in the observed stripe patterns.
FIG 5C shows a three-dimensional model 21 indicating the characteristics of the surface 13 obtained through triangulation relations. Optical 3d measuring systems use the method of triangulation to determine the spatial dimensions and geometry of a surface 13. The projection centers of the projector 11 and the camera 14 and the pattern on the surface 13 define a spatial triangle. By determining the angles between the pixels of the projector 12 and the image and the basis, the intersection point and thus the 3Dcordinate can be calculated using triangular relations. The map as discussed herein corresponds to the 3D model 21. The map can be used to determine the correspondence between the projector pixels 12

and the image pixels 16. The displacement of any stripe can be directly converted into 3D co-ordinates using triangulation.
By projection of a sequence of pixels 12 in a binary coded stripe pattern 22, the subsequent brightness level for each surface pixel is forming a unique binary word revealing the individual stripe number. Further, every point on the triangulated stripe is located, as the extents of the stripe correspond to extend of the pixel on the projector 11. . Following every stripe gives the correspondence of every pixel, which can be further used to establish corresponding points between the pixels 12 in the projector 12 and the image 16 captured by the camera 14.
FIG 6A-6B shows an exemplary illustration of an embodiment of the invention. The images 16 illustrate how each object on the surface 13 has its own individual structured light pattern. FIG 6A shows an image 16 of a surface 13 impinged with a set of projector pixels 12 in a normal pattern. The surface 13 includes multiple objects such as a resistor 27 and a thermocole pyramid 28. Here, the resistor 27 is of low contrast and the thermocole pyramid 28 is of high contrast.
The objective is to increase the contrast of the resistor 27 to make it more visible while retaining the contrast of the rest of the surface 13. The projector pixels 12 which are projected to the resistor 27 and captured in the corresponding image pixels 16 are located. Further, the brightest of the pixels 12 are adjusted to increase the contrast of the resistor 26.
The processing means 17 iterates the projector 11 and an optimal brightness distribution scheme is obtained at the end

of the iteration. The surface 13 is then illuminated with optimal brightness and the camera 14 produces the image 16 as shown in FIG 6B, where the resistor 26 appears brighter with better contrast for inspection purposes. The iterative scheme adapts the brightness of the pixels 12 using the feedback provided by the camera 14 and projects pixels 12 of different brightness to different objects. Here, the brightness of the projector pixels 12 corresponding to the resistor 26 is increased so as to increase the contrast of the resistor,26 for improving surface characteristics.
FIG 7A-7D shows an exemplary illustration of another embodiment of the invention. In FIG 7A, the image 16 of the surface 13 including a thermocole pyramid 28 illuminated with a set of projector pixels 12 in a uniform pattern. When a normal light is impinged to the surface, the brightness of pixels reaching various parts of the thermocole pyramid 28 is different.
FIG. 7B herein illustrates a graph 30 which shows the distribution of brightness on the surface of the thermocole pyramid 28. The graph 30 indicates that the brightness is less in the left and right sides of the pyramid 28 whereas the brightness is more towards the center. This variation is caused by the shape of.the pyramid 28 with different depths. The distribution of brightness herein corresponds to the projector pixels 12 incident on the surface 13.
For optimal surface inspection, the surface of the pyramid 28 should be uniformly illuminated. The processing means 17 adjust the distribution of pixels 12 at the projector 11 such that the pixels 12 projected from the projector 11 are focused to different regions of the pyramid 28. The

brightness of the pixels 12 are adjusted so as to provide a uniform illumination on the pyramid 28.
The pattern of projector pixels 12 required for optimal illumination of the pyramid 28 is generated in response to the identified features and/or surface texture of the pyramid 28. The brightness of pixels 12 is made uniform over the thermocole pyramid 28 using the corrected pixel pattern as shown in FIG. 7B.
The graph 30 according to FIG 7D shows the distribution of brightness along the surface of the pyramid 28 is substantially uniform. The projector pixels 12 are calibrated to project pixels of adapted brightness corresponding to the geometry of the pyramid 28.
FIG 8 illustrates a flow diagram depicting a method of increasing the contrast of the image 16 according to an -embodiment herein. A set- of projector pixels 12 are projected to the surface 13 to be inspected for surface defects in step 81. The projector pixels 12 incident on the surface 13 and a corresponding image 16 of the surface 13 is captured by a camera 14 at step 82. Further, the image is analyzed to identify a portion 18 of the surface with a low contrast in the image 16 in step 83. In step 84, projector pixels 12 which are projected to a part of the surface 13 and are captured in the identified portion 18 of the image 16 is located.
For obtaining information of the projector pixels 12 to be projected to the surface 13 for optimal illumination, every pixel 12 of the projector 11 should be mapped onto the corresponding pixel 16 captured by the camera 14. To perform the pixel mapping, a map 21 is created by projecting the set

of projector pixels 12 in a predefined geometric pattern to the surface 13. A corresponding image 16 of the surface is obtained and then calculates the pixel correspondence between the projector 11 and the camera 14.
Further, in step 85, the brightness of the located projector pixels 12 are adjusted such that the contrast of the identified portion 18 is increased. The brightness of the -pixels 12 at the projector 11 is adjusted so as to; subsequently increase or decrease the brightness of the pixels to bring the brightness to the desired level.
The embodiment describes a method to improve the quality of captured image 15 by virtualizing a high-resolution projector 11. The projected pixels 12 as seen in the captured image 15 can be thought as the intersection of two sets of equally spaced parallel lines which are approximately orthogonal. By calculating the orientation and frequency of these lines, and •■■•then taking the intersection of- them, we identify the location of the projected pixels 12 in the captured image. Before that the captured image 16 is converted into a gray level and local contrast is normalized.
The embodiment herein can be used anywhere where large scaled projections of high quality are to be produced at low cost and without consuming much time. For example, the embodiment herein finds extensive application in precision shape measurement for production control for e.g. in turbine blades, in reverse engineering to obtain precision CAD data from existing objects, volume measurement of objects, precision structure measurement of grinded surfaces, radius determination of cutting tool blades, precision measurement of planarity, skin surface measurement for cosmetics and medicine, body shape measurement, forensic inspections, Skin

surface measurement for cosmetics and medicine, wrinkle measurement on cloth and leather, measurement of topography of solar cells, and the like.
The embodiment herein includes software for controlling both the hardware and for enabling the computer or microprocessor to interact with an end user or other mechanism utilizing the results of the present embodiment. Such software may include device drivers, operating systems, user applications and the like.
While this invention has been described in detail with reference to certain preferred embodiments, it should be appreciated that the present invention is not limited to those precise embodiments. Rather, in view of the present disclosure which describes the current best mode for practicing the invention, many modifications and variations would present themselves, to those of skill in the art without departing from the scope and spirit of this invention. The scope of the invention is, therefore, indicated by the following claims rather than by the foregoing description. All changes, modifications, and variations coming within the meaning and range of equivalency of the claims are to be considered within their scope.

Patent claims WE CLAIM:
1. A method of increasing a contrast of an image (15) of a
surface (13), said method comprising steps of:
- projecting a set of projector pixels (12) to the surface (13);
- capturing the image (15) of the surface (13);
- identifying a portion (18) in the image (15) with a low contrast (13) corresponding to a part of the surface (13);
- locating the projector pixels (12) which are projected to the part of the surface (13) and are captured in the identified portion (18) of the image (15); and
- adjusting a brightness of the located projector pixels (12) such that the contrast of the identified portion (18) is increased.

2. The method according to claim 1, wherein a map is used to locate the projector pixels (12) corresponding to the identified portion (18) of the image (15). -
3. The method according to claim 2, wherein the map is obtained by:

- projecting the set of projector pixels (12) in a predefined geometric pattern (22) to the surface (13);
- obtaining an image (15) of the surface (13);
- identifying the location of the projector pixels (12) of the predefined geometric pattern (22) in the image (15); and
- correlating the predefined geometric pattern of pixels (12) projected to the surface (13) and the identified in the image
(15) to create the map.
4. The method according to claim 3, wherein the predefined
geometric pattern (22) is a striped pattern (23).

5. The method according to claim 3, wherein the predefined geometric pattern (22) is a checkerboard pattern (24).
6. The method according to any of the claims 2 to 5, wherein the mapping is performed for the identified portion (18) of a predefined size.
7. The method according to any of the claims 1 to 6, wherein adjusting the brightness includes increasing the brightness of the identified portion (18) if the brightest image pixels, (16) in the identified portion (18) are darker than an identified threshold..
8. The method according to any of the claims 1 to 7, wherein adjusting the brightness of the identified portion (18) includes decreasing the brightness of the identified portion if the darkest image pixels (16) in the identified portion (18) are brighter than the identified threshold.
9. The method according to any of the claims 1 to 8, wherein the method is used to detect surface defects of the surface (13).
10. A system (10) for increasing a contrast of an image (16)
of a surface (13), the system (1.0) comprising:
- a projector (11) to project a set of projector pixels (12) to the surface (13);
- a camera (14) for capturing an image (15) of the surface (13); and
- a processing means (17) adapted to:
- identify a portion (18) in the image (15) with a low contrast corresponding to a part of the surface (13);

- locate the projector pixels (12) which are projected to the part of the surface (13) and are captured in the identified portion (18) of the image (15); and
- control the projector (11) to adjust a brightness of the located projector pixels (12) such that the contrast of the identified portion (18) is increased.

11. A system according to claim 10, wherein the processing means (17) is adapted to perform the method according to any of the claims 2 to 10.
12. The system according to claim .10 or 11, wherein the projector (11) includes at least one of a micromirror device
(19) and a liquid crystal display to project the set of projector pixels (12) to the surface (13).
13. The system according to claim 10 or 11, wherein the
camera (14) is at least one of a CCD camera and a CMOS camera.

A method and system for increasing a contrast of an image of surface comprising steps of projecting a set of projector pixels to the surface using a projector, capturing an image of the surface using a camera, and a processing means for identifying a portion in the image with a low contrast corresponding to a part of the surface, locating the projector pixels which are projected to the part of the surface and are captured in the identified portion of the . image and adjusting a brightness of the located projector pixels such that the contrast of the identified portion is increased is disclosed. A map is used to locate the projector pixels corresponding to the identified portion of the image.