The present technology relates to a surgical system, an image processing device, and an image processing method capable of obtaining the posture of an eye more firmly and with a high degree of accuracy. An image information acquisition unit acquires a tomographic image that is a cross-sectional image taken in a direction substantially parallel to an eye axis direction of an eye of a patient. An image recognition unit recognizes each part of the eye on the tomographic image by means of image recognition on the basis of the tomographic image. The control unit calculates the posture of the eye on the basis of the result of recognition of each part of the eye. In this way, by recognizing each part of the eye on the basis of the tomographic image and calculating the posture of the eye on the basis of the result of recognition, the posture of the eye can be obtained more firmly and with a high degree of accuracy. The present technology can be applied to a surgical system.
