Abstract: To meet the demand for improving the imaging quality of payloads on the new generation ocean color observation satellite, optimization design and control measures were conducted for key factors affecting imaging quality, such as signal-to-noise ratio (SNR), scanning mechanism stability, and stray light suppression, etc. The SNR was improved by means of enlarging the light entrance aperture, increasing the scanning pixel count and integration time, optimizing the optical path, and adjusting the time-delay integration (TDI) stages and gain levels. The linearity of the scanning mechanism's stability was enhanced by measures like constant-speed scanning, slow start/slow stop design, and improvement of the angle measurement resolution. Stray light was suppressed by a combined structure of light shield and light baffle. The drift angle was corrected by real-time yaw control in the Y-direction. Test results show that the SNR of the Chinese ocean color and temperature scanner (COCTS) reaches up to 1 760, the linearity of the scanning mechanism is 0.24%, the stray light coefficient is ≤1%; the dynamic modulation transfer function (MTF) of the multi-spectral bands of the coastal zone imager (CZI) is ≥0.20, and the MTF of the panchromatic band is ≥0.15. The on-orbit image quality of various payloads is excellent, providing high-quality data support for marine resource development, environmental monitoring, etc.