Multi-stage progressive target recognition training system and method based on electroencephalogram

Abstract

The invention discloses a multi-stage progressive target recognition training system and method based on electroencephalogram. The system comprises an electroencephalogram cap, a signal amplifier, a computer, a training system interface and a P300 decoding unit, professionals help trainees wear the electroencephalogram cap, open the training system interface and enter different training interfacesaccording to needs, after training is started, the trainees watch a display interface, the interface randomly displays a target picture and a non-target picture. The electroencephalogram cap collectselectroencephalogram signals of a trainee at the moment, the electroencephalogram signals are amplified by the signal amplifier and then transmitted to the P300 decoding unit to be decoded, and afterdecoding succeeds, a decoding result is used for judging whether the trainee recognizes a target picture or not, then the recognition rate is calculated, and the recognition capacity of the trainee is evaluated. According to the method, convenience is provided for man-machine mixed tiny target image recognition work, the recognition capability of workers is evaluated from the perspective of P300electroencephalogram data, and a basis is provided for selecting excellent artificial recognition professionals.

Description

technical field [0001] The present invention relates to a multi-stage progressive target recognition training system and method based on EEG. On the one hand, it provides a new method for personnel engaged in human-machine mixed micro-target image recognition. The training platform based on P300 EEG enables the staff to pass Multi-stage progressive image recognition training gradually familiarizes with the target features and adapts to the target playback frequency, thereby improving the accuracy of manual recognition. On the other hand, it provides a method to evaluate the staff's recognition ability from the perspective of objective analysis of P300 EEG data, so as to provide a basis for selecting excellent artificial recognition professionals. Background technique [0002] With the continuous progress and development of artificial intelligence, target recognition technology has been widely used in many fields, such as vehicle automatic driving obstacle recognition, sea fl...

AI Technical Summary

Problems solved by technology

However, for the detection of some easily confused targets with small volume and inconspicuous outlines, the existing technology is still difficult to accurately identify them, so it is necessary to further use artificial recognition methods for targets that are difficult for computers to identify, that is, human-machine mixed tiny targets Image recognition method

At present, the main shortcomings are: due to the large amount of detection data in the real environment, the real-time requirements are high, and for untrained personnel, they are not familiar with the target features that need to be recognized, and at the same time they cannot quickly adapt to the established playback frequency. It is difficult to correctly identify the required target objects, and there is a lack of a training platform that can provide staff with corresponding training and gradually improve the recognition ability; in addition, due to the innate individual differences of the staff in terms of target recognition ability, it is often necessary to select a machine with certain inherent advantages. of the staff complete the target recognition task, so a system that can effectively evaluate the staff's personal ability is urgently needed

Images