The invention belongs to the field of brain-computer interface devices, and particularly relates to an animal-emotion electroencephalogram analysis device. The animal-emotion electroencephalogram analysis device comprises an animal electroencephalogram collection module, a preprocessing module, a feature classification module and an analysis-result output module. By means of the animal-emotion electroencephalogram analysis device, eye electrical noise, drifting noise and high-frequency noise in single-electrode original electroencephalogram signals can be removed, and according to feature classification based on machine learning, current-emotion state values of animals are obtained, broadcasted and displayed. Electroencephalogram is collected through a claw-type dry electrode, and can penetrate through hair to be in contact with the cortex, and is combined with the preprocessing module, and it can be guaranteed that the signal to noise ratio is high, and wearing is more convenient; classification is completed through machine learning, and the effectiveness is guaranteed; meanwhile, the time resolution ratio can be increased, and finer emotional state changes can be caught. The animal-emotion electroencephalogram analysis device has the advantages of being fast and convenient to wear and high in sensitivity and accuracy rate, and is suitable for application and popularization.