基于深度学习的目标检测研究

Research on object detection based on deep learning

目标检测是利用对图像信息开展高效、精准定位来进行相关信息的识别、预定义,进而实现对物体类别的判断,但是基于深度学习的目标检测需要开展相关优化与技术改进方面的内容仍较多。文章主要分析和研究基于深度学习的目标检测过程中关于主流目标、双阶段目标、单阶段目标方面的技术提升的相关策略,以期推动计算机技术中的视觉、模式识别等技术得到进一步发展。

Target detection is the use of image information to carry out efficient and accurate positioning to identify,pre-define,and then achieve the object category judgment,but the depth learning-based object detection needs to carry out the relevant optimization and technical improvement content is still more.This paper mainly analyzes and studies the mainstream target,two-stage target,but stage target technology promotion strategies in the process of deep learning target detection,in order to promote the further development of vision,pattern recognition and other technologies in computer technology.