Target detection method based on evolutionary neural network under constraint condition

Abstract

The invention discloses a target detection method based on an evolutionary neural network under a constraint condition, and the method comprises the steps: constructing a plurality of structural blocks and a population composed of a plurality of individuals, and carrying out the coding of each individual through a variable-length coding mode, thereby completing the initialization of the population; performing training updating on each individual according to the training data set; evaluating the individuals on the verification data set, and calculating the accuracy and complexity of the individuals to obtain the fitness of the individuals; according to a preset constraint quantity, adjusting the individual fitness by using a constraint control method, and adjusting the individual framework of which the accuracy exceeds a threshold value; selecting male parents from the population according to the adjusted fitness, generating first-level offspring through male parent crossing, and generating second-level offspring through probabilistic variation of the first-level offspring; and selecting the parent, the first-level filial generation and the second-level filial generation to generate a new population, and performing iterative evolution. According to the invention, a light-weight structure unit is designed, a constraint method is utilized, and an optimized target detection result is achieved without artificial experience.

Description

technical field [0001] The invention relates to the field of digital image processing, in particular to a target detection method based on an evolutionary neural network under constraint conditions. Background technique [0002] Object detection is a research hotspot in the fields of computer vision, machine learning, artificial intelligence, etc. It has a wide range of applications in areas such as intelligent video surveillance, robot environment perception, and large-scale image retrieval. However, object detection is still a challenging task due to various deformations, pose changes, and environmental factors such as background lighting and angles in actual scenes. With the continuous development of deep learning technology, the performance of object detection based on deep learning has been greatly improved. At present, target detection mainly uses image processing and deep learning methods to locate the target of interest in the image, accurately distinguishes the cat...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the above-mentioned deficiencies in the prior art, the present invention provides a target detection method based on an evolutionary neural network under constraints, which solves the problem that the current target detection model has a large scale and cannot be directly used in mobile terminals and embedded devices with limited resources. And the problem that the lightweight neural network in the existing technology is very dependent on human experience

Images