Abstract: Techniques for image segmentation can include receiving image data of an image including a background and a face of a person in a foreground, and determining a respective a priori probability of a head-shoulder foreground pixel appearing per pixel of the plurality of pixels, according to a positioning result of a plurality of exterior contour points of the face. The techniques can also include selecting foreground and background pixels of the plurality of pixels, according to at least the a priori probabilities, and determining respective color likelihood probabilities of the foreground and the background, according to color feature vectors of the selected pixels. The techniques can also include determining respective posteriori probabilities of at least part of the foreground and at least part of the background, according to the a priori probabilities and the respective color likelihood probabilities.

Abstract: A method for a device to adjust a skin color of an image includes: identifying a skin color region of the image; performing a statistical calculation on original color data of pixels in the skin color region to obtain an original mean value and an original standard deviation of the original color data of the pixels in the skin color region; selecting a preset skin color model from one or more preset skin color models each representing one skin color type, according to the original mean value and a preset mean value of each of the one or more preset skin color models; determining target color data according to the original color data, the original mean value, the original standard deviation, and the preset mean value and a preset standard deviation of the selected skin color model; and adjusting the skin color region according to the target color data.

Abstract: Techniques for image segmentation can include receiving image data of an image including a background and a face of a person in a foreground, and determining a respective a priori probability of a head-shoulder foreground pixel appearing per pixel of the plurality of pixels, according to a positioning result of a plurality of exterior contour points of the face. The techniques can also include selecting foreground and background pixels of the plurality of pixels, according to at least the a priori probabilities, and determining respective color likelihood probabilities of the foreground and the background, according to color feature vectors of the selected pixels. The techniques can also include determining respective posteriori probabilities of at least part of the foreground and at least part of the background, according to the a priori probabilities and the respective color likelihood probabilities.