Re-interpolation with edge detection for extracting an aging pattern for AMOLED displays
A method that includes an initial uniform pixel measurement and interpolation followed by an edge detection algorithm to recognize the areas that contribute mostly to the estimation error due to the interpolation. The pixels on the detected edges and around their vicinity are also measured, and an aging pattern of the entire display is obtained by re-interpolating the entire measured set of data for the initially measured pixels as well as the pixels around the detected edges. The estimation error is reduced particularly in the presence of aging patterns having highly spatially correlated areas with distinctive edges.
1. An image compensation method for a display having pixels that are aging due to shifts in one or more aging characteristics of the pixels, each of the pixels comprising a light emitting device, the method comprising:
measuring, using a measurement circuit, an aging characteristic of each pixel of a spaced apart subset of the pixels of the display to produce a set of initial pixel measurements;
determining localized maximums of a gradient magnitude defined by the set of initial pixel measurements;
defining a second subset of pixels of the display with use of the determined localized maximums of the gradient magnitude;
measuring, using the measurement circuit, the aging characteristic of each pixel of the second subset of pixels of the display to produce a second set of pixel measurements;
producing an aging pattern based on the set of initial pixel measurements and the second set of pixel measurements;
storing the aging pattern in a memory device;
generating, with use of the aging pattern, an estimation matrix corresponding to a pixel resolution of the display for improving an image uniformity of the light emitting devices across the display.
2. The method of claim 1 , wherein the spaced subset of the pixels of the display comprises a spaced uniform distribution of pixels.
3. The method of claim 2 , wherein the spaced uniform distribution of pixels comprises a spaced rectilinear distribution of pixels.
4. The method claim 3 , wherein the spaced rectilinear distribution of pixels comprises at least some but not all pixels of the display using a down-sampling rate of KV×KH, such that the aging characteristic of at least every KVth pixel is measured along a column and the aging characteristic of at least every KHth pixel is measured along a row.
5. The method of claim 1 , wherein producing the aging pattern based on the set of initial pixel measurements and the second set of pixel measurements comprises interpolating a combination of the set of initial pixel measurements and the second set of pixel measurements.
6. The method of claim 5 wherein determining localized maximums of the gradient magnitude comprises interpolating the set of initial pixel measurements to produce an initial aging pattern and using an edge detection algorithm on the initial aging pattern.
7. The method of claim 1 , wherein defining the second subset of pixels comprises selecting one or more pixels within a local vicinity of each of the determined localized maximums of the gradient magnitude.
8. The method of claim 1 , wherein each entry in the estimation matrix corresponds to a respective pixel of the display, and wherein each value of each entry in the estimation matrix corresponds to an amount by which the respective pixel is aging such that a corresponding compensation value is applied to increase the programmed brightness for the pixel to compensate for the amount of aging.
9. The method of claim 1 , wherein the aging characteristic is related to at least one of: a shift in a threshold voltage of a drive transistor that drives the light emitting device in each of the pixels; a change in a voltage across the light-emitting device in each of the pixels; a change in a drive current of the drive transistor needed to cause the light-emitting device in each of the pixels to emit a programmed brightness; and a change in a current of the light emitting devices needed to emit a programmed brightness.
10. The method of claim 1 , wherein interpolating the combination of the set of initial pixel measurements and the second set of pixel measurements generates estimated aging characteristics for pixels of the display other than those for which set of initial pixel measurements and the second set of pixel measurements were made.
11. The method of claim 1 , wherein determining localized maximums of the gradient magnitude comprises performing a Canny edge detection algorithm on the magnitudes defined by the set of initial pixel measurements.