Gradient domain high dynamic range compression
read more
Citations
Guided Image Filtering
Computer Vision: Algorithms and Applications
Poisson image editing
Poisson surface reconstruction
Guided image filtering
References
Scale-space and edge detection using anisotropic diffusion
Numerical Recipes in C: The Art of Scientific Computing
Lightness and Retinex Theory
Adaptive histogram equalization and its variations
Related Papers (5)
Frequently Asked Questions (19)
Q2. What are the future works mentioned in the paper "Gradient domain high dynamic range compression" ?
Future work will concentrate on the many different exciting possible applications of the construction of an image from modified gradient fields. In addition, the authors would like to extend their work so as to in- 7Exposures courtesy of Max Lyons, c© 2001 Max Lyons.
Q3. What is the main advantage of TRCs?
The main advantage of TRCs lies in their simplicity and computational efficiency: once a mapping has been determined, the image may be mapped very quickly, e.g., using lookup tables.
Q4. How does the dynamic range compression process work?
Dynamic range compression is achieved by scaling down the smoothest version, and then adding back the differences between successive levels in the hierarchy, which contain details removed by the simplification process.
Q5. How many parameters are used to achieve an optimal result?
In addition, the method is controlled by no less than 8 parameters, so achieving an optimal result occasionally requires quite a bit of trial-and-error.
Q6. How many exposure times were used to take these photographs?
These photographs were taken using a digital camera with exposure times ranging from 1/1000 to 1/4 of a second (at f/8) from inside a lobby of a building facing glass doors leading into a sunlit inner courtyard.
Q7. How do they eliminate the notorious halo artifacts?
In order to eradicate the notorious halo artifacts Tumblin and Turk [1999] introduce the low curvature image simplifier (LCIS) hierarchical decomposition of an image.
Q8. What boundary conditions are natural in the case of divG?
In their case, the most natural choice appears to be the Neumann boundary conditions ∇I ·n = 0 (the derivative in the direction normal to the boundary is zero).
Q9. How long does it take to compress a high-resolution image?
the LCIS hierarchy construction is computationally intensive, so compressing a high-resolution image takes a substantial amount of time.
Q10. What is the way to capture the color and texture of the stone tiles?
The color and texture of the stone tiles just outside the door are best captured in the middle image, while the green color and the texture of the ficus plant leaves becomes visible only in the very last image in the sequence.
Q11. What is the common way to acquire a panoramic image?
A popular way to acquire a panoramic image is to scan a scene using a video camera and then construct a mosaic from the video frames.
Q12. What is the effect of the LCIS image?
The LCIS image (right) exhibits a grainy texture in smooth areas, and appears to slightly overemphasize edges, resulting in an “embossed”, non-photorealistic appearance.
Q13. How does the method attenuate large gradients?
Their method attenuates large gradients and then constructs a low dynamic range image by solving a Poisson equation on the modified gradient field.
Q14. What is the purpose of this paper?
In this paper, the authors present a new technique for high dynamic range compression that enables HDR images, such as the one described in the previous paragraph, to be displayed on LDR devices.
Q15. What is the way to obtain an HDR panorama?
If the authors let the camera’s auto-exposure control set the correct exposure for each frame, each scene element is imaged at multiple aperture settings and the authors can construct an HDR as in [Debevec and Malik 1997].
Q16. What is the solution to the gradient attenuation problem?
It is important to note that although the computation of the gradient attenuation function is done in a multi-resolution fashion, ultimately only the gradients at the finest resolution are manipulated, thus avoiding halo artifacts that typically arise when different resolution levels are manipulated separately.
Q17. What is the effect of attenuating the low frequencies in homomorphic filtering?
As a result, attenuating only the low frequencies in homomorphic filtering may give rise to strong “halo” artifacts around strong abrupt changes in illuminance, while Horn’s method incorrectly interprets sharp shadows as changes in reflectance.
Q18. What is the efficient method to search the space of all 2D potential functions?
In their method the authors employ a more direct and more efficient approach: search the space of all 2D potential functions for a function The authorwhose gradient is the closest to G in the leastsquares sense.
Q19. How do the authors change the magnitudes of the gradients?
in order to avoid introducing spatial distortions into the image, the authors change only the magnitudes of the gradients, while keeping their directions unchanged.