Video target detection method based on machine learning

Abstract

The invention discloses a video target detection method based on machine learning. The method comprises the steps that (1) for an input video, an SSD target detection algorithm is adopted to obtain ato-be-tracked target detection box, and a bounding-box is marked on an image to determine a tracking target; (2) two tracking methods are adopted for each frame of the input video, wherein one tracking method is a light stream tracking algorithm, a tracking point of the next frame is predicted according to a probability, and the tracking point of the next frame is precisely determined through a Euclidean distance and a set threshold value; and the other tracking method is to adopt a full-convolutional neural network and extract high-layer features and low-layer features in the neural network for separate convolution, finally the features are fused into a feature graph through a classifier, and therefore the tracking point of the next frame is precisely determined; and (3) HOG features of the light stream tracking result and the full-convolutional neural network tracking result are extracted, validity discrimination is performed on the two results through a support vector machine (SVM),and the target position of the next frame is determined finally.

Description

technical field [0001] The present invention relates to the field of computer vision, and more specifically, relates to a video target detection method based on machine learning. Background technique [0002] The development of science and technology has made camera equipment popular, and a large amount of image data and video data have emerged as the times require. Among them, video data has also received extensive attention. Many fields need to use target detection and tracking, such as surveillance video, drones, etc. tracking etc. In these applications, a target candidate box is usually given first, and then tracked, and the target to be tracked cannot be automatically identified. In particular, if the target appears in multiple videos, it is not suitable for humans to find and track the target in a large number of videos, and these tracking systems are also not suitable. Therefore, solving the automatic identification of tracking targets can greatly improve the applic...

AI Technical Summary

Problems solved by technology

[0004] The above method is partly for static image target detection, partly for video target detection, and the target detection method used in video target tracking is also to improve the tracking effect, and does not realize automatic recognition and tracking, so it cannot be applied to tracking automatic target recognition and tracking

Images