Facial nerve paralysis detection system in desensitization state based on thermal imaging and depth map

Abstract

The invention relates to the technical field of computer vision, in particular to a facial nerve paralysis detection system in a desensitization state based on thermal imaging and a depth map. The system comprises a facial acquisition module, a facial correction module, a symmetry axis division module and a facial anomaly detection module which are connected in sequence, the problems that the detection result error is large and data desensitization cannot be realized due to division of a facial symmetry axis by an existing facial anomaly detection system are solved. According to the invention,data desensitization is realized by collecting the thermodynamic image and the depth image of the patient, and meanwhile, in a data desensitization state, sensitive information of the face of the user is further protected by utilizing the sparse depth value of the tile, so that the user experience is improved; besides, through face correction and symmetry axis division of each row of tiles, errors caused by face deflection and face symmetry axis division are greatly reduced, and the accuracy and reliability of a detection result are improved.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a facial nerve paralysis detection system in a desensitized state based on thermal imaging and a depth map. Background technique [0002] At present, disease detection systems based on mobile terminals generally need to collect RGB images, and RGB images contain sensitive information such as the patient's environment and facial details; however, some patients do not want the detection system to collect information such as the current indoor environment and facial details, such as: Facial stubble. [0003] In addition, facial anomaly detection based on depth maps usually uses the longitudinal central axis of the image or the line connecting specific feature points as the facial symmetry axis. However, this facial symmetry axis is not ideally collinear, and patients with facial paralysis sometimes have their mouths shifted as a whole. Therefore, Using the longitudinal cent...

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a facial nerve paralysis detection system under desensitization state based on thermal imaging and depth map, and solves the technical problem that the division of facial symmetry axes by the existing facial abnormality detection system will cause large errors in detection results and cannot be realized. data desensitization

Images