Split edge enhancement architecture
View Patent ↗A system and method for enhancing the detail edges and transitions in an input video signal. This enhancement may be accomplished by enhancing small detail edges before up-scaling and enhancing large amplitude transitions after up-scaling. For example, detail edge enhancement (detail EE) may be used to enhance the fine details of an input video signal. An edge map may be used to prevent enhancing the large edges and accompanying mosquito noise with the detail enhancement. Noise may additionally be removed from the signal. After the fine details are enhanced, the signal may be up-scaled. Luminance transition improvement (LTI) or chrominance transition improvement (CTI) may be used to enhance the large transitions of the input video signal post scaler.
1. A method comprising:
receiving a video signal that includes fine detail edges and large edges, wherein the large edges have a greater transition in the image relative to the fine detail edges;
differentiating, based on the greater transition, the large edges in the video signal from the fine detail edges in video signal;
enhancing the fine detail edges in the video signal without enhancing the large edges in the video signal to yield a first detail-enhanced signal;
scaling the first detail-enhanced signal to convert the first detail-enhanced signal from a first display resolution to a second display resolution;
scaling the video signal including both the fine detail edges and the large edges to convert the video signal from the first display resolution to the second display resolution;
enhancing the large edges in the scaled video signal without enhancing the fine detail edges in the scaled video signal to yield a second detail-enhanced signal; and
combining the first detail-enhanced signal with the second detail-enhanced signal.
2. The method of claim 1 , wherein enhancing the fine detail edges in the video signal without enhancing the large edges in the video signal includes:
high-pass filtering and gaining of the video signal to enhance the video signal; and
applying a large-edge mask to the enhanced video signal to mask out enhancements of the large edges in the video signal.
3. The method of claim 1 , wherein enhancing the fine detail edges in the video signal without enhancing the large edges in the video signal includes:
high-pass filtering and gaining horizontal edges of the video signal to yield a horizontal-edge-enhanced signal; and
separately high-pass filtering and gaining vertical edges of the video signal to yield a vertical-edge-enhanced signal; and
combining the horizontal-edge-enhanced signal with the vertical-edge-enhanced signal to yield the first detail-enhanced signal.
4. The method of claim 1 , wherein the first detail-enhanced signal and the second detail-enhanced signal are combined by adding the first detail-enhanced signal to the second detail-enhanced signal to yield a final output signal.
5. The method of claim 4 , wherein enhancing the large edges in the scaled video signal comprises applying a luminance transition improvement (LTI) to the scaled video signal including both the fine detail edges and the large edges.
6. The method of claim 4 , wherein enhancing the large edges in the scaled video signal comprises applying a chrominance transition improvement (CTI) to the scaled video signal including both the fine detail edges and the large edges.
7. The method of claim 4 , wherein enhancing the large edges in the scaled video signal includes applying, to the scaled video signal including both the fine detail edges and the large edges, a luminance bias that is a positive function of a second derivative of the scaled video signal including both the fine detail edges and the large edges.
8. The method of claim 7 , wherein the second derivative is in the horizontal direction.
9. The method of claim 7 , further comprising:
determining a minimum pixel value and a maximum pixel value,
wherein enhancing the large edges in the scaled video signal includes biasing the scaled video signal including both the fine detail edges and the large edges i) towards the maximum pixel value if the second derivative is a positive number, and ii) towards the minimum pixel value if the second derivative is a negative number.
10. The method of claim 1 , wherein differentiating the large edges in the video signal from the fine detail edges in video signal includes
evaluating pixels in the video signal to identify large edges in the video signal,
expanding each identified large edge horizontally and vertically, and
selecting additional pixels, that surround the respective identified large edge, to be included in the respective edge.
11. A method comprising:
receiving a video signal;
enhancing the video signal to yield a detail-enhanced signal by
high-pass filtering and gaining horizontal edges of the video signal to yield a horizontal-edges-enhanced signal,
high-pass filtering and gaining vertical edges of the video signal to yield a vertical-edge-enhanced signal, and
combining the horizontal-edge-enhanced signal with the vertical-edge-enhanced signal to yield the detail-enhanced signal, wherein the enhancing of the video signal enhances fine detail edges in the video signal without enhancing large edges in the video signal;
scaling the detail-enhanced signal to yield a scaled detail-enhanced signal, wherein the detail-enhanced signal has a first display resolution and the scaled detail-enhanced signal has a second display resolution;
scaling the video signal to yield a scaled input signal, wherein the video signal has the first display resolution and the scaled input signal has the second display resolution;
enhancing preferentially the large edges of the scaled input signal to yield a large-edge-enhanced scaled signal; and
adding the large-edge-enhanced scaled signal to the scaled detail-enhanced signal to yield a final output signal.
12. The method. of claim 11 , wherein the enhancing of the fine detail edges in the video signal without enhancing the large edges in the video signal includes:
applying a large-edge mask to the detail-enhanced signal to mask out enhancements of the large edges in the detail-enhanced signal.
13. A system comprising:
a large-edge map configured to (i) receive a video signal that includes fine detail edges and large edges, wherein the large edges have a greater transition in the image relative to the fine detail edges and (ii) to differentiate, based on the greater transition, the large edges of the video signal from the fine detail edges in the video signal;
a detail calculator configured to enhance the fine detail edges in the video signal without enhancing the large edges in the video signal to yield a first detail-enhanced signal;
a video scaler configured to scale the first detail-enhanced signal, wherein the scaling of the first detail-enhanced signal converts the first detail-enhanced signal from a first display resolution to a second display resolution;
a second video scaler for scaling the video signal to yield a scaled signal, wherein the video signal has the first display resolution and the scaled signal has the second display resolution;
a large-edge enhancer for enhancing the large edges of the scaled signal to yield a second detail-enhanced signal; and
an adder for adding the second detail-enhanced signal to the scaled first detail-enhanced signal to yield a final output signal.
14. The system of claim 13 , wherein the detail calculator includes a high-pass filter and a gain device to enhance the video signal.
15. The system of claim 13 , further comprising an adder that combines the enhanced signal with a mask output of the large-edge map to mask out enhancements of the large edges.
16. The system of claim 13 , wherein the detail calculator includes high-pass filters and gains horizontal edges of the input video signal to yield a horizontal-edge-enhanced signal, and separately high-pass filters and gains vertical edges of the video signal to yield a vertical-edge-enhanced signal, and combines the horizontal-edge-enhanced signal with the vertical-edge-enhanced signal.