Continuous and rapid visual demonstration non-target picture optimal sorting method based on color moments

Abstract

The invention discloses a continuous rapid visual demonstration non-target picture optimal sorting method based on color moments, and mainly solves the problem that event-related potential (ERP) is induced by errors due to the fact that similarity of adjacent pictures of non-target pictures is too low during RSVP presentation in existing continuous rapid visual demonstration. According to the implementation scheme, the method comprises the steps of converting an RGB picture into an HSV picture, and extracting color moments of pixels on all channels of the HSV picture to serve as features; converting each channel pixel value of one HSV picture to generate a simple conversion picture, manually sorting the simple conversion pictures, sorting according to the similarity of the simple conversion pictures, and determining each channel weight parameter in combination with two sorting results; and for all non-target pictures, calculating picture similarity by utilizing the weight parameters ofeach channel, and finishing sorting. According to the method, optimal sorting of RSVP non-target pictures can be realized, the quality of ERP can be effectively improved, and the method can be used for a brain-computer interface.

Description

technical field [0001] The invention belongs to the technical field of electronic information, and mainly relates to a method for sorting non-target pictures in continuous and fast visual demonstration, which can be used for brain-computer interface. Background technique [0002] With the continuous advancement of social information technology, the problem of information overload is becoming more and more serious. As image and video data repositories are growing at an exponential rate, the size, diversity, and potential sparsity of "objects of interest" in these data repositories pose difficulties for efficient retrieval of objects. Continuous Rapid Visual Demonstration RSVP is a BCI paradigm derived from the combination of the human visual system and the event-related potential ERP of the cerebral cortex under the environment of continuous development of brain-computer interface BCI technology in recent years. It is often used to help professionals, such as satellite images...

AI Technical Summary

Problems solved by technology

The problem lies in ignoring the fact that the similarity between adjacent non-target pictures is too low, which leads to the ERP signal being easily induced by mistake, which affects the classification results.

However, the problem with this method is that Phash only takes the low-frequency part of the picture in order to tolerate some deformation of the picture, which causes the problem that the features cannot capture the details of the picture, which makes the processing effect of pure color pictures or approximate pure color pictures very poor

Adjacent non-target images of pure color with excessively large color differences are easily identified as similar images by the Phash algorithm, causing ERP to be induced, which in turn may determine the non-target image as the target image, resulting in false alarms

Images