The article presents retinex theory, which explains how the human visual system perceives lightness and color despite variations in illumination. The theory proposes that the visual system uses three independent cone systems, each sensitive to long-, middle-, and short-wavelength light. These systems generate separate images of the world in terms of lightness, which are not mixed but compared to produce color sensations. The key insight is that lightness is independent of illumination, as demonstrated by experiments where changing the flux of light did not alter perceived color. The authors describe a method to calculate lightness values that are biologically correlated with reflectance. They also show that color sensations are independent of the flux-wavelength distribution of light, as evidenced by experiments where changing the light flux did not change perceived color. The theory suggests that the visual system uses a processing mechanism that is instantaneously independent of the wavelength-luminance composition of light. This mechanism is not controlled by time-dependent processes, such as changes in visual pigments. The experiments also show that the visual system can determine reflectance without direct comparison to a standard. The authors propose a model of the visual system that uses sequential products of ratios of luminances at adjacent points to determine reflectance. This model is independent of illumination and is consistent with the observed independence of color sensation from flux-wavelength distribution. The theory is supported by experiments with a colored Mondrian, where changes in illumination did not alter perceived color. The authors conclude that the visual system uses a processing mechanism that is not merely independent, but instantaneously independent of the wavelength-luminance composition of light. This mechanism is not controlled by time-dependent processes, such as changes in visual pigments. The theory is supported by experiments with a colored Mondrian, where changes in illumination did not alter perceived color. The authors conclude that the visual system uses a processing mechanism that is not merely independent, but instantaneously independent of the wavelength-luminance composition of light.The article presents retinex theory, which explains how the human visual system perceives lightness and color despite variations in illumination. The theory proposes that the visual system uses three independent cone systems, each sensitive to long-, middle-, and short-wavelength light. These systems generate separate images of the world in terms of lightness, which are not mixed but compared to produce color sensations. The key insight is that lightness is independent of illumination, as demonstrated by experiments where changing the flux of light did not alter perceived color. The authors describe a method to calculate lightness values that are biologically correlated with reflectance. They also show that color sensations are independent of the flux-wavelength distribution of light, as evidenced by experiments where changing the light flux did not change perceived color. The theory suggests that the visual system uses a processing mechanism that is instantaneously independent of the wavelength-luminance composition of light. This mechanism is not controlled by time-dependent processes, such as changes in visual pigments. The experiments also show that the visual system can determine reflectance without direct comparison to a standard. The authors propose a model of the visual system that uses sequential products of ratios of luminances at adjacent points to determine reflectance. This model is independent of illumination and is consistent with the observed independence of color sensation from flux-wavelength distribution. The theory is supported by experiments with a colored Mondrian, where changes in illumination did not alter perceived color. The authors conclude that the visual system uses a processing mechanism that is not merely independent, but instantaneously independent of the wavelength-luminance composition of light. This mechanism is not controlled by time-dependent processes, such as changes in visual pigments. The theory is supported by experiments with a colored Mondrian, where changes in illumination did not alter perceived color. The authors conclude that the visual system uses a processing mechanism that is not merely independent, but instantaneously independent of the wavelength-luminance composition of light.