Image processing method and terminal

Abstract

The embodiment of the invention provides an image processing method and terminal. The method comprises the steps that a positive sample set and a negative sample set are acquired; the positive sample set and the negative sample set are smoothed to obtain a target positive sample set and a target negative sample set; the target positive sample set and the target negative sample set are subjected to integral channel feature extraction to obtain X first target feature sets; pixel difference value feature extraction is performed on each of the X first target feature sets to obtain X second target feature sets, wherein each of the X second target feature sets contains multiple pixel difference value features; and at least one boosting decision tree is adopted to classify the X second target feature sets to obtain a target classifier. Through the embodiment, the classifier with good feature screening capability and high calculation speed can be obtained on the premise that performance is not lowered.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image processing method and a terminal. Background technique [0002] In the existing technology, face detection is a very basic research direction in the field of vision, and its application range is very wide, such as security, smart business, robotics, human-computer interaction and other fields. Since the human face is a non-rigid object, it is easily disturbed by the external environment, such as weather, light, expression, posture, camera distortion and other factors, the difficulty of detection is obvious. Commonly used face detection methods include early algorithms based on face geometry, template matching algorithms, machine learning-related algorithms, and popular deep learning algorithms in recent years. [0003] Among the existing face detectors, the feature expression of a class of sub-detectors is relatively complex, such as the 10 channel feat...

AI Technical Summary

Problems solved by technology

[0003] Among the existing face detectors, the feature expression of a class of sub-detectors is relatively complex, such as the 10 channel features in the integral channel graph and the HOG feature in DPM. Although they show good performance in face detection, the calculation is too large. Complicated, with a large amount of feature calculation and large redundant information, it is difficult to meet the requirements of real-time applications

Images