NEW, Junior Level Developer from Pakistan Needed. I need a robust, well-documented Python solution that can scan batches of images, detect visible defects, and output concise reports I can plug straight into my quality-control pipeline. The exact defect categories and image sources will be finalized together—think scratches, discolorations, cracks on anything from standard photographs to X-ray or microscopic captures—but the core requirement is a clean, modular codebase that lets me switch models or add new classes with minimal refactoring. Here’s how I picture the workflow: the script ingests a folder (or live feed) of images, runs preprocessing, applies the chosen detection model, flags each defect with bounding boxes or masks, and finally exports a CSV/JSON summary along with annotated images. I’m comfortable installing extra libraries, so you’re free to leverage OpenCV, scikit-image, TensorFlow, PyTorch, or any combination that gets reliable results fast. Deliverables • Python code with clear, inline comments • A requirements.txt or environment.yml for painless setup • Brief README covering execution steps and key parameters • Sample run on a small test set that demonstrates defect detection and report generation Acceptance criteria • Script executes from the command line in one step • Defect detection accuracy and false-positive rate are benchmarked in your sample run • Code structure allows me to add new defect classes without rewriting core logic If you’ve built similar inspection tools—or have pretrained models I can fine-tune—let me know. I’m ready to start as soon as we confirm the technical approach.