基于 TensorFlow 的简单的垃圾分类系统 A Simple Garbage Classification SystemBased on TensorFlow 内容摘要内容摘要：垃圾分类是近段时间来，全国范围内的热点问题，它是需要落实到我们每一个人身上来。但是由于范围太广且是最近才开展的工作，所以还是有“混合垃圾”出现。人工智能机器学习也是今年来快速发展的技术，那么利用 TensorFlow 去简单进行垃圾分类其实也是非常不错的选择。本文就使用 TensorFlow 中比较常见且使用非常广泛的卷积神经网络来实现智能垃圾分类的目标，卷积神经网络是现今图像分类的比较好的解决方案。TensorFlow 扎根于Python 中，并且还是跨平台的，兼容性方面自然不必多讲，主要是其 整 个 的 生 态 是 十 分 健 全 的 ， 在 Python 当 中 的 不 少 库 如keras，pytorch，random,OpenCV 等，都能够对 TensorFlow 有着不小的助力。本文主要通过卷积神经网络去训练处能够识别分为四大类垃圾：“可回收的垃圾，有害的垃圾，厨余的垃圾，其他的垃圾”的模型，然后通过该模型来完成对垃圾图像的识别和分类，帮助我们去完成垃圾分类这个主要的任务。关键词： TensorFlow垃圾分类 图像识别 ResNet 模型机器学习卷积神经网络PythonAbstractAbstract：Garbage classification is a hot issue in the country recently, which needs to be implemented in every one of us. But because the scope is so wide and the work is so recent, there is still mixed garbage. Artificial intelligence machine learning is also a rapidly developing technology this year, so TensorFlow is actually a very good choice for simple garbage classification. Here I want to discuss how to make garbage classification intelligent, so I think of tensorflow, and then use convolutional neural network to realize it. Convolutional neural network is a better solution for image classification nowadays. Here I want to discuss how to use convolution neural network in tensorflow to realize the goal of garbage classification intelligently. Convolutional neural network is a better solution for image classification n...
