LDRAW based renders of LEGO bricks moving on a conveyor belt with extracted models

The set contains renders of LEGO bricks moving on a white conveyor belt. The images were prepared for training neural network for recognition of LEGO bricks. For each brick starting position, alignment and color was selected (simulating the brick falling down on the conveyour belt) and than 10 images was created while the brick was moved across the conveyor belt. Afterwards empty frames, with no brick visible, were removed from the set. The images were saved in JPEG format. All images were generated using Blender (https://www.blender.org/) tool and were based on the 3D models from LDraw (https://www.ldraw.org/) brick library. The bricks were than extracted from the original images using OpenCV edge detection algorithms.
The colors for the LEGO bricks were selected from the following list [color (code)]: White (0xffffff), Brick Yellow (0xD9BB7B), Nougat (0xD67240), Bright Red (0xff0000), Bright Blue (0x0000ff), Bright Yellow (0xFfff00), Black (0x000000), Dark Green (0x009900), Bright Green (0x00cc00), Dark Orange (0xA83D15), Medium Blue (0x478CC6), Bright Orange (0xff6600), Bright Bluish Green (0x059D9E), Bright Yellowish-Green (0x95B90B), Bright Reddish Violet (0x990066), Sand Blue (0x5E748C), Sand Yellow (0x8D7452), Earth Blue (0x002541), Earth Green (0x003300), Sand Green (0x5F8265), Dark Red (0x80081B), Flame Yellowish Orange (0xF49B00), Reddish Brown (0x5B1C0C), Medium Stone Grey (0x9C9291), Dark Stone Grey (0x4C5156), Light Stone Grey (0xE4E4DA), Light Royal Blue (0x87C0EA), Bright Purple (0xDE378B), Light Purple (0xEE9DC3), Cool Yellow (0xFFFF99), Medium Lilac (0x2C1577), Light Nougat (0xF5C189), Dark Brown (0x300F06), Medium Nougat (0xAA7D55), Dark Azur (0x469bc3), Medium Azur (0x68c3e2), Aqua (0xd3f2ea), Medium Lavender (0xa06eb9), Lavender (0xcda4de), White Glow (0xf5f3d7), Spring Yellowish Green (0xe2f99a), Olive Green (0x77774E), Medium-Yellowish Green (0x96B93B).
The original folder contains the renders themselves, the cropped_opencv directory contains only bricks extracted from th erenders. In both cases the images were placed in a folder named after the LEGO brick code (as read from LDraw). The files naming convetion is as follows
brickID_colour_sequenceNumber_timestamp.jpg
where brickID is the LEGO brick id number as read from LDraw, color is the name of the selected brick color, sequenceNumber is the integer from 0 to 9 indicating the number of the image in the sequence and timestamp is UNIX time representation in milliseconds of the image creation time.
Sample images are presented below.

本数据集包含在白色传送带上运动的乐高（LEGO）积木渲染图。本数据集的图像用于训练用于乐高积木识别的神经网络。针对每一块积木，我们设置其初始位置、朝向与颜色（模拟积木从上方掉落至传送带的过程），并在积木沿传送带移动的过程中生成10张图像。随后，我们移除了数据集中无积木可见的空帧。所有图像均以JPEG格式存储，通过Blender（https://www.blender.org/）工具生成，并基于LDraw（https://www.ldraw.org/）积木库中的3D模型构建。后续我们使用OpenCV边缘检测算法从原始图像中提取出积木区域。

本次选用的乐高积木颜色来自以下列表[颜色（色码）]：白色（0xffffff）、砖黄色（0xD9BB7B）、肉色（0xD67240）、亮红色（0xff0000）、亮蓝色（0x0000ff）、亮黄色（0xFFFF00）、黑色（0x000000）、深绿色（0x009900）、亮绿色（0x00cc00）、深橙色（0xA83D15）、中蓝色（0x478CC6）、亮橙色（0xff6600）、亮蓝绿色（0x059D9E）、亮黄绿色（0x95B90B）、亮红紫色（0x990066）、沙蓝色（0x5E748C）、沙黄色（0x8D7452）、土蓝色（0x002541）、土绿色（0x003300）、沙绿色（0x5F8265）、深红色（0x80081B）、火焰黄橙色（0xF49B00）、红棕色（0x5B1C0C）、中石灰色（0x9C9291）、深石灰色（0x4C5156）、浅石灰色（0xE4E4DA）、浅宝蓝色（0x87C0EA）、亮紫色（0xDE378B）、浅紫色（0xEE9DC3）、冷黄色（0xFFFF99）、中淡紫色（0x2C1577）、浅肉色（0xF5C189）、深棕色（0x300F06）、中肉色（0xAA7D55）、深天蓝色（0x469bc3）、中天蓝色（0x68c3e2）、水绿色（0xd3f2ea）、中薰衣草色（0xa06eb9）、薰衣草色（0xcda4de）、夜光白（0xf5f3d7）、春黄绿色（0xe2f99a）、橄榄绿（0x77774E）、中黄绿色（0x96B93B）。

数据集的原始文件夹存储所有渲染原图，`cropped_opencv`目录仅包含经OpenCV从原图中裁剪提取的积木图像。两种场景下的图像均被存入以乐高积木编号（取自LDraw库）命名的文件夹中。文件命名规则如下：
"brickID_colour_sequenceNumber_timestamp.jpg"
其中，`brickID`为取自LDraw库的乐高积木编号，`colour`为所选积木的颜色名称，`sequenceNumber`为0至9的整数，代表该序列中的图像序号，`timestamp`为图像生成时刻的毫秒级UNIX时间戳。下文展示了示例图像。