Huang S.-CJaw D.-WChen B.-HSY-YEN KUO2021-09-022021-09-02202121686750https://www.scopus.com/inward/record.uri?eid=2-s2.0-85074852414&doi=10.1109%2fTETC.2019.2943231&partnerID=40&md5=a451b01a9662ccc6a1113739880113c5https://scholars.lib.ntu.edu.tw/handle/123456789/581164Sufficient illumination enables the observer to identify the true colors of an object captured by an imaging sensor, whereas dim illumination restricts proper observation. This paper proposes a single image enhancement algorithm based on human visual system. First, the image is transformed by the proposed Luminance Perception procedure from the spatial domain into the proposed Brightness Domain, where the image can be assessed as to whether or not it has sufficient illumination. Further, if the image is determined to have insufficient illumination, the assessed range of insufficient illumination and sharpness is used to restore the image's true illumination and textures to best fulfill the requirements for human visual perception. Our experiments show that the enhanced images generated by the proposed algorithm exhibit more brightness and appreciably better contrast than those generated by other state-of-the-art image enhancement algorithms that use images captured under various illuminant levels in various kinds of natural scenes. ? 2013 IEEE.Computer graphics; Graphics processing unit; Lighting; Luminance; Program processors; Textures; Histograms; Human visual perception; Human Visual System; Imaging sensors; Observers; Spatial domains; State of the art; Visual systems; Image enhancementjournal article10.1109/TETC.2019.29432312-s2.0-85074852414