I论 文 题 目：基于深度学习的图像目标检测算法的研究及应用摘 要人类所接触的信息大都是视觉信息，自古以来，人们凭借看到的信息模仿学习从而进步乃至创造，即便是到了现如今科技的时代，图像和视频这些由高科技带来的视觉信息也同样是人们愿意从中找出兴趣快乐甚至投身研究的首选，而目标检测便是人们为了利用好这些视觉信息而应运而生的前沿技术，它不断被突破技术难关，攻坚克难，已经成为了当前计算机视觉识别领域的中流砥柱，它的技术并不是单一的，而是涵盖着涉及人工智能、识别分类等很多方面的高端研究成果，并大范围应用在智能化系统系列、农产品检测、物品分类等领域。所谓图像目标检测，就是对输入的图像进行目标锁定，范围圈定，类别确定，而当面对视频也就是动态的图像时，也可以做到精准定位，把所需要的图像识别出来。本文对图像目标检测算法进行了研究，提出新的角度想法，想要使用卷积神经网络来对目标检测算法中传统目标检测方法的一些不好的地方进行改进，同时也对比分析了算法实验。关键词：目标检测；卷积神经网络；深度学习；分类回归IIABSTRACTMost of the information that humans are exposed to is visual information. Since ancient times, people have imitated learning and made progress or even created by the information they see. Even in the era of technology, the visual information brought by high technology such as images and videos is also People are willing to find happiness and even the first choice for research, and target detection is the cutting-edge technology that people came up with in order to make good use of these visual information. It has been constantly broken through technical difficulties and overcome difficulties. It has become the current computer vision. The mainstay in the field of identification, its technology is not single, but covers high-end research results involving many aspects of artificial intelligence, identification and classification,and is widely used in intelligent system series, agricultural product detection, item classification and other...
