ISP Image Processing Algorithm Engineer
Job Responsibilities
- Responsible for developing endoscopic ISP imaging and processing algorithms based on NVIDIA Holoscan and other platforms, and cooperating with FPGA engineers to debug and optimize image parameters;
- Continue to track the latest research results at home and abroad in the ISP field, participate in the construction and improvement of image quality testing standards;
- Write relevant technical documents and technical patents according to the work results.
Job Requirements
- Master degree or above in computer, automation, biomedical engineering, optics or related field;
- Proficient in programming languages such as C++, Python or MatLab, familiar with the principle of parallel acceleration, with FPGA, NVIDIA CUDA, OpenGL development experience is preferred, familiar with deep learning image processing algorithm is preferred;
- Have actual ISP development experience, be proficient in related algorithms and processing processes of ISP pipeline, and be able to independently complete algorithm development and maintainer of multiple modules (such as Demosaic, DNR, bad point compensation, Gamma calibration, gamut space conversion, sharpness, 3A, HDR, WDR, etc.) is preferred. Special consideration for those with AI-ISP experience;
- Familiar with sensor imaging principle and understand the impact of various parts of the camera, such as Sensor, lens, filter, ISP, etc., on image quality. ISP debugging experience of mainstream platforms such as HiSilicon and Anba is preferred;
- Familiar with the analysis method and calculation principle of objective indicators of image evaluation (such as dynamic range, MTF, signal-to-noise ratio, color reproduction, low illumination and other performance indicators), master Imatest test tool or other test platform, familiar with medical image evaluation criteria is preferred.
- Medical endoscope ISP development experience is preferred.