Abstract: A non-contact facial blood pressure measurement method based on 3D CNN is disclosed, which belongs to the technical field of computer vision. The method includes the following steps. S110: collecting an actual face video sample and training a blood pressure prediction model based on face images using 3D CNN neural network. S120: obtaining a face video in real time through a HD camera. S130: recognizing face key points in the face video obtained in S120 through dlib face recognition model, selecting a face region of interest, and extracting face images from the region. S140: performing a wavelet transform operation on the face images extracted in S130 to remove noise. S150: inputting seven consecutive frames of the face images into the 3D CNN blood pressure prediction model trained in S110 to obtain a blood pressure value of the measured person. The disclosure realizes non-contact facial blood pressure measurement.

Abstract: The present invention discloses a food detection and identification method based on deep learning, which realizes food positioning and identification by a deep convolutional network. The method comprises: firstly, training a general multi target positioning network and a classification network by using food pictures; secondly, inputting the results of the positioning network into the classification network; finally, providing a classification result by the classification network. The method uses two deep convolutional networks with different functions to respectively detect and identify the food, which can effectively reduce the labeling cost of the food and improve the accuracy of positioning and identification.

Abstract: The present invention discloses a food detection and identification method based on deep learning, which realizes food positioning and identification by a deep convolutional network. The method comprises: firstly, training a general multi target positioning network and a classification network by using food pictures; secondly, inputting the results of the positioning network into the classification network; finally, providing a classification result by the classification network. The method uses two deep convolutional networks with different functions to respectively detect and identify the, food, which can effectively reduce the labeling cost of the food and improve the accuracy of positioning and identification.