A road image target detection method based on an attentional mechanism, relating to the field of machine learning technology in the field of artificial intelligence, for application in the field of smart cities, and comprising acquiring a road image to undergo detection (202); inputting the road image to undergo detection into a trained integrated detection framework (204) and attentional weight values corresponding to each target detection framework; by means of the target detection frameworks, performing target detection processing on the road image to undergo detection, and outputting obtained pedestrian detection results (206); on the basis of the attentional weight values, combining the pedestrian detection values to obtain a target detection result of the integrated detection framework (208). Further provided are a road image target detection apparatus based on an attentional mechanism, a computer device, and a storage medium. In addition, the invention further relates to blockchain technology. The described method is able to solve the technical problem in the prior art that it is difficult for target detection frameworks to implement indicators set in a targeted manner to perform target detection.
展开▼