Target detection method based on synthetic data set

Abstract

The invention discloses a target detection method based on a synthetic data set. The method includes: adding a real environment picture as a background map to the three-dimensional model of the to-be-detected target in 3ds MAX software so as to establish a three-dimensional scene, rendering and generating a required number of synthetic images, and automatically completing the marking of the imagecategory and the marking box so as to complete the construction of a composite data set; using the synthetic data set as a training set to train the target detection network; and carrying out target detection after training is completed. By using the method, the labeled data set of any target can be quickly obtained at low cost, and the problems that the real data set is high in labeling cost andreal data cannot be obtained in a specific scene are solved. Furthermore, the designed target detection network is added with the SOMConv layer, so that the identification capability of the network onreal data can be improved.

Description

technical field [0001] The invention belongs to the field of target detection, and relates to the construction of a synthetic data set in target detection and a target detection method suitable for the synthetic data set. Background technique [0002] In the target detection task, the target detection algorithm based on deep learning has completely surpassed the target detection algorithm based on the non-deep learning method and has become the mainstream target detection algorithm. But deep learning-based object detection algorithms rely on large-scale annotated datasets. [0003] Object detection dataset construction is a key technology in object detection tasks. The sample size and types of the target detection data set largely determine the effect of the target detection algorithm. During the construction of the target detection data set, a large number of target pictures to be detected will be obtained, and then the obtained pictures will be labeled with data. During ...

AI Technical Summary

Problems solved by technology

The data labeling work of data set construction is relatively cumbersome and costly. Most of the research on target detection algorithms is carried out on public data sets. Public data sets provide a large number of labeled pictures, which saves research costs. There are fewer types of data sets that cannot meet specific needs

At the same time, in some specific scenarios, such as dangerous scenes in automatic driving and some non-public targets, it is difficult to obtain target images at this time, and the cost of data acquisition is relatively high

Images