Expression detecting device and method

Abstract

The invention relates to expression detecting device and method. The expression detecting device comprises a gray-scale image generating unit, a human face edge detecting unit, a dynamic complexion capturing unit, a human face contour generating unit and an expression detecting unit. The gray-scale image generating unit is used for generating a gray-scale image according to an original image. Thehuman face edge detecting unit is used for outputting a human face edge detecting result according to the gray-scale image. The dynamic complexion capturing unit is used for generating a dynamic complexion capturing result according to the original image and generating a human face and background segmenting result according to the dynamic complexion capturing result. The human face contour generating unit is used for outputting a human face contour according to the gray-scale image, the human face edge detecting result and the human face and background segmenting result. The expression detecting unit is used for outputting an expression detecting result according to the human face contour.

Description

technical field [0001] The present invention relates to an expression detection device and an expression detection method thereof, and in particular to an expression detection device with a low computational load and an expression detection method thereof. Background technique [0002] In life, people often express their inner emotions through facial expressions. Emotional expression areas mainly include: mouth, eyes, eyebrows and cheeks. When people express their inner emotions, they can express a state of mind by only changing some local features of the face (such as raising the corners of the mouth). With the rapid development of technology, people further expect expression recognition to be applied to electronic devices, so as to greatly improve the convenience of application. [0003] Smile detection is one of the important items of facial expression detection, and its processing method can be divided into two parts: face feature detection technology and classifier. ...

AI Technical Summary

Problems solved by technology

[0004] However, the traditional facial expression detection technology not only has a huge amount of calculation, but also is not easy to be used in embedded platforms (such as digital cameras)

In addition, the traditional facial expression detection technology is easily affected by the light source. When the brightness is uneven, it directly affects the correctness of the expression detection results.

Images