Hardware Implementation and Verification of the Sensor Defective Pixels Correction Algorithm

摘要

Algorithms for correction of defective pixels of sensor raw data are considered. A modification of the directed average sum algorithm is proposed. Implemented module includes multiple stages conveyor. Two parameters that can influence on module’s and conveyor’s performance: size of color component and number of pixels per clock. A numerical study of the quality of image correction by the selected algorithm depending on the module parameters was carried out, as well as a comparison of the result with correction by the average method. A verification environment has been developed, and the module has been verified via UVM (IEEE 1800.2-2017 Universal Verification Methodology).