Parallel Gamma Correction Circuit for Multi-Drop Inkjet Heads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing circuits for inkjet heads face challenges in achieving high processing speed for gamma correction while maintaining low power consumption and circuit scale, particularly due to limitations in hardware multipliers and memory requirements for nozzle variations.
Innovation Solution
An image processing circuit that performs parallel gamma correction calculations for gray level values 1 to Z using a calculator and selector, eliminating the need for hardware multipliers and allowing implementation in a low-cost FPGA, thereby increasing processing speed without increasing power consumption or circuit scale.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a hardware multiplier is used in the image processing circuit to perform gamma correction calculations, then the calculation processing speed is improved, but the circuit scale and power consumption increase
Solution Approach 1:
The gamma correction calculation is segmented into two independent stages: first, the gain table memory stores pre-calculated gain values for each gray level and nozzle; second, the selector circuit retrieves and applies the appropriate gain value to the input signal. This segmentation eliminates the need for a hardware multiplier, reducing circuit scale while maintaining acceptable processing speed through pre-computation.
Solution Approach 2:
The gain values for gamma correction are pre-calculated and stored in the gain table memory before the actual image processing occurs. By performing the multiplication operation in advance and storing the results, the circuit avoids real-time multiplication during image processing, thereby eliminating the need for a hardware multiplier and reducing both circuit scale and power consumption.
2Manufacturing precision
If gamma correction is performed per nozzle to correct ejection characteristic variations, then the print image density uniformity is improved, but the processing complexity and memory requirements increase
Solution Approach 1:
The gain table memory stores separate gain values for each nozzle, allowing each nozzle to have its own customized correction characteristics. This local quality approach enables precise correction of individual nozzle ejection variations without requiring complex per-nozzle processing circuits, as the correction data is simply stored and retrieved based on the nozzle identifier.
Solution Approach 2:
The gain table memory structure serves multiple functions: it stores correction data for all nozzles, allows retrieval of specific nozzle correction values, and enables uniform processing of all nozzles through the same selector circuit. This multi-functionality reduces overall system complexity compared to having separate processing circuits for each nozzle.
3Speed
If a high-cost FPGA with high-speed circuit is used to increase operation clock rate, then the processing speed is improved, but the power consumption and cost increase
Solution Approach 1:
The invention uses a low-cost FPGA implementation by replacing the hardware multiplier with a gain table memory and selector circuit. While the operation clock rate may be lower than what a high-cost FPGA could achieve, the solution provides adequate processing speed for the application while dramatically reducing power consumption and cost, accepting a trade-off that is acceptable for the intended use case.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
A calculator performs in parallel gamma correction calculations on gray level values 1 to Z depending on each of target nozzles for each of target pixels containing gray level values other than zero of pixels in image data to be printed by a multi-drop inkjet head including nozzles and having a maximum droplet number Z per pixel, the nozzles of the multi-drop inkjet head including the target nozzles for ink ejection on the respective target pixels. A selector selects a calculation result corresponding to a gray level value of the target pixel from calculation results of the gamma correction calculations performed on the gray level values 1 to Z by the calculator, for each of the target pixels.