Adaptive generation of Q-table2 for improved image quality
View Patent ↗Brightness correction and image enhancement are derived to compensate for changes in captured image data by hardware scanners and the like by statistically measuring the captured data and generating or selecting a compensation quantization table to be substituted in image data encoded for compression. Three techniques of compensation usable singly or in combination are provided which may be used in combination with techniques for dynamic range reduction for extreme compression. Thus, improved images which may be enhanced in any desired manner, for example, to conform to an established image standard, may be derived from data having sufficient compression to allow economically acceptable long-term storage in high-speed access storage media even though the data captured is degraded in dynamic range, brightness and the like.
1. A method of compressing image data including steps of
capturing image data to derive captured image data,
statistically measuring said captured image data to develop an image data metric,
quantizing and encoding said captured image data to derive encoded and compressed image data,
substituting a quantization table in accordance with said image data metric or data representing said image data metric into said encoded and compressed image data, and
computing an average of a fraction of pixels in said captured image data having highest brightness.
2. A method as recited in claim 1 wherein said fraction equals approximately 90% of pixels in an image.
3. A method as recited in claim 1 wherein said fraction is equal to or greater than 5% of pixels in an image.
4. A method of compressing image data including steps of
capturing image data to derive captured image data,
statistically measuring said captured image data to develop an image data metric,
quantizing and encoding said captured image data to derive encoded and compressed image data, and
substituting a quantization table in accordance with said image data metric or data representing said image data metric into said encoded and compressed image data,
wherein said step of substituting a quantization table includes a further step of
computing or selecting a quantization table having values which increase brightness values approximately proportionally relative to a desired maximum brightness and captured image data of maximum brightness.
5. A method as recited in claim 4 wherein said computing or selecting step provides a quantization table having values which are approximately multiples of values of a quantization table used for performing said quantizing and encoding step.
6. A method as recited in claim 4 wherein said step of substituting a quantization table includes further steps of
determining darkening of an image represented in said captured image data, and
providing a compensating offset in orthogonal transform coefficients.
7. A method of compressing image data including steps of
capturing image data to derive captured image data,
statistically measuring said captured image data to develop an image data metric,
quantizing and encoding said captured image data to derive encoded and compressed image data, substituting a quantization table in accordance with said image data metric or data representing said image data metric into said encoded and compressed image data, and
compressing dynamic range of said captured image data prior to said step of quantizing and encoding,
wherein said step of substituting a quantization table includes further step of
computing or selecting a quantization table having values which approximate a product of values which increase brightness values approximately proportionally relative to a desired maximum brightness and captured image data of maximum brightness and values which approximately compensate for said step of compressing dynamic range.
8. A method as recited in claim 7 wherein said computing or selecting step provides a quantization table having values which are values that are products of values approximately multiples of values of a quantization table used for performing said quantizing and encoding step and one or more values which approximately compensate for dynamic range reduction.
9. A method as recited in claim 7 wherein said step of substituting a quantization table includes further steps of
determining darkening of an image represented in said captured image data, and
providing a compensating offset in orthogonal transform coefficients.
10. A method of compressing image data including steps of
capturing image data to derive captured image data,
statistically measuring said captured image data to develop an image data metric,
quantizing and encoding said captured image data to derive encoded and compressed image data, and
substituting a quantization table in accordance with said image data metric or data representing said image data metric into said encoded and compressed image data,
wherein said step of substituting a quantization table includes further steps of
determining darkening of an image represented in said captured image data, and
providing a compensating offset in orthogonal transform coefficients.
11. Apparatus for compressing image data including
a sensor for capturing image data to derive captured image data,
a data processor for statistically measuring said captured image data to develop an image data metric, quantizing and encoding said captured image data to derive encoded and compressed image data, and substituting a quantization table in accordance with said image data metric or data representing said image data metric into said encoded and compressed image data, and
means for storing or transmitting said encoded and compressed image data including at least one of said quantization table and said data representing said metric,
means for compressing dynamic range of said captured image data prior to said step of quantizing and encoding, and
means for decoding and displaying said encoded and compressed image data in accordance with said quantization table substituted into said encoded and compressed image data or a quantization table derived from said metric.