A command-line image-processing program that locates faces with an OpenCV Haar cascade, displays bounding boxes and writes detection information to JSON.
This is face detection, not identity recognition. The source supports JPEG, PNG and a GIF-reading path. JsonCpp is bundled third-party code: its embedded authorship and license notices remain authoritative.
C++17 · OpenCV · giflib · JsonCpp
Image path → image loading → Haar cascade detection → rectangles
├─ JSON output
└─ annotated display
The cascade location is hard-coded for a Linux-style installation. GIF support introduces an additional native dependency; the GUI display also requires a graphical environment. These are reproducibility constraints, not evidence of a packaged cross-platform release.
The supplied Makefile selects GNU C++17 and OpenCV libraries. Before building, configure OpenCV headers/libraries, link giflib (-lgif) and make the Haar cascade available at HAAR_CASCADE_PATH in FaceDetector.hpp. The committed Makefile does not include giflib linkage, and the cascade XML is not bundled. Once these prerequisites are resolved, the program takes an image path as its first argument: ./FaceDetector path/to/image.jpg. A clean build was not verified during this documentation review.
Christian Vladimir Sucuzhanay Arévalo
Data & AI Solutions Architect | AWS Data Architecture | Generative AI & Amazon Bedrock | Big Data | Former University Lecturer
Entity Home · Technical Portfolio · LinkedIn · AWS Builder · GitHub
Build. Explain. Teach. Share.