ARC-100
收藏资源简介:
ARC 100数据集是作为名为ARC (自动零售结帐) 的零售结帐原型系统的一部分收集的。它由巴基斯坦拉合尔的100种常见零售商品的31,000 640美元 \ times 480美元RGB图像组成。每个零售商品都有310图像,这些图像由Logitech C310网络摄像头在各种逻辑方向 (黑色,哑光传送带上) 捕获,在LED灯条照明的木制引擎盖框架下 (亮度设置为约 $ 70lx $)。在建议的设置中,在输入卷积神经网络进行识别之前,对图像进行预处理和标准化。 链接: 圆弧纸 ARC-100数据集
The ARC 100 dataset was collected as part of a retail checkout prototype system named ARC (Automatic Retail Checkout). It comprises 31,000 640×480 RGB images of 100 common retail goods gathered in Lahore, Pakistan. Each retail item has 310 images captured by a Logitech C310 webcam under various logical orientations on a black matte conveyor belt, within a wooden hood frame illuminated by LED light bars, with the brightness set to approximately 70 lux. In the proposed setup, images are preprocessed and normalized before being fed into a convolutional neural network for recognition. Links: curved paper, ARC-100 dataset




