Target detection method based on Faster-RCNN reinforcement learning

Abstract

The invention discloses a target detection method based on Faster-RCNN reinforcement learning and relates to the field of image processing. The method comprises acquiring an image to be tested; introducing the image to be tested into a Faster-RCNN; modifying the network structure of the CNN in the Faster-RCNN to replace a convolution module in the network structure of the last scale with a hourglass module; extracting the feature of the image to be tested by the CNN to generate feature mapping maps; importing the last feature mapping map into an RPN; vectorizing a feature mapping map corresponding to a candidate region selected by the RPN and then classifying the feature mapping map by a classifier to obtain a detection result. The method modifies the network structure of the CNN, replacesthe ordinary convolution module in the deep network with the hourglass module, performs reinforcement learning on the semantic information carried by the deep features extracted by the deep CNN, hierarchically highlights the semantic information of an object, and reduces the false negatives and false positives to a certain extent.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a target detection method based on Faster-RCNN reinforcement learning. Background technique [0002] Object detection is widely used in the fields of pedestrian detection, intelligent assisted driving, intelligent monitoring, flame and smoke detection, and intelligent robots. Although object detection technology is developing rapidly, there are still many problems. Illumination, occlusion, background confusion, and scale problems have always been problems. Difficulties in object detection. [0003] The deep convolutional neural network performs very well on the target detection task, mainly due to the large sample and its complex form, and the depth makes the model have a strong nonlinear expression ability. Usually, after the deep convolutional neural network is built, end-to-end training is carried out. Although the convolutional neural network visualization technology can be u...

AI Technical Summary

Problems solved by technology

[0002] Object detection is widely used in the fields of pedestrian detection, intelligent assisted driving, intelligent monitoring, flame and smoke detection, and intelligent robots. Although object detection technology is developing rapidly, there are still many problems. Illumination, occlusion, background confusion, and scale problems have always been problems. Difficulties in target detection

Usually, after the deep convolutional neural network is built, end-to-end training is carried out. Although the convolutional neural network visualization technology can be used to observe the pros and cons of network training, but at present this is only a basis for judging whether the network is converged , the visualized deep features extracted by the convolutional neural network contain rich semantic information, but currently there is no follow-up processing on these visualized features. If the semantic information can be relearned and refined, it will be useful for subsequent image processing tasks. helpful

Images