Upper Body 2,332 images. Various image size about : [100~300] x [100~300]. 

opencv + aws lambda source code + material
details: www.marearts.com/AWS-Lambda-OpenCV-d406790e9bfa4512a544dc26c57235f8

Extracting all image objects in PDF file. python version code. More detail is here : www.marearts.com/Image-Objects-Extracting-in-PDF-d3f1f1f5d3574660987074e04119ccdd

Natural Korean car number plate images for OCR. Details : www.marearts.com/Dataset-for-Korean-car-license-plate-OCR-f5747d21157343f2919e35a807c25b33

Python source code to detect face and eyes, nose, lips and making mosaic.
Details: www.marearts.com/Rotation-invariant-Face-Eye-Nose-Lips-Detector-and-Mosaic-471e51e051744e1c91a9aa0a66a345ba

python source code to convert yolo2labelme, labelme2yolo format. details: www.marearts.com/Labelme2Yolo-Yolo2Labelme-3564dd886ac64b5499d6f2784a8a4be8

MareArts Realtime Stitching SDK Detail : www.marearts.com/Realtime-Stitching-SDK-ba67e56550a14688835f291f891af4ac

6 types of artificial generated Korean car license dataset- Every type has 5,000 images so total about 30,000 images, 1.83GB- It was augmented with various effects, check more detail : www.marearts.com/Artificial-data-of-Korean-car-license-plates-3847a00c841e4542b5d2a3e3157cc25a

Korean Car license plate image generator python source code, detail : www.marearts.com/Korea-Car-License-Plate-Generator-990e115a3dd44a1eb86c2c0c6685f2f1

18,175 car images in various background with label txt file, 5.47 GB.coordinate information x1, y1, x2, y2(left, top, right, bottom), detail:www.marearts.com/Dataset-for-car-license-plate-detection-6f1e623b60c94e54ac6f55bbd5a6a917

Compiled tesseract 5.0 executable binary file for Amazon Linux(Cent OS) and AWS Lambda

Human Head 18,305 imagesVarious Image Size about [100~10] x [100~10]

8,590 pedestrian images, Size is various about [100~200] x [300~400], Different weather (day, night, rainy)

Cube shape images from DIDSON sonar equipment - 24x24, 2089

Frontal Face 19x19, 2429 jpg images

MNIST handwritten digit jpeg image and origin binary dataset.24x24 and 70,000 images. More details here : www.marearts.com/Data-Warehouse-931708779aea4dc5be444dbfff547aa5?p=5d60021839304c53aa16b21d60a01672&pm=s

PDF structure analysis python source code (extract json, image, structure as per page)

Tea Time! ☕️ Computer Vision #3: Recall, Precision, F1 and ROC, AUC  lecture material, 2 version PPT(origin, drawing)

OpenCV + Webapp Lecture source code and video url to see all contents

OpenCV Lecture class materials(pdf), source code and images

The dataset contains frontal, full profile 90° and half profile 45° face images. The face image was obtained from 22 people. The images were taken in various lighting locations. 100x100 dimension & total 35,325 images.

C++ source code to draw dash line using openCV

Image processing python code for mosaic effect. The source code has following functions : superpixel, subdiv, delaunay, voronoi

480x360, left right image from bumblebee stereo camera.658 jpeg.Unpaved, paved roads.

Face image dataset, 24x24, 19832 images