A 3-month consultancy to design, prototype, build, and validate a production-ready in-app whole-plot image capture feature for the Ona Android application, enabling breeding teams to capture plot images without leaving the app.
Responsibilities
- Review the current Ona capture flow, codebase, and previous panorama development work
- Build quick prototypes for the most promising plot-image capture options and test them on target devices
- Compare capture approaches based on capture time, stitching or counting success rate, image quality, phone memory, battery and heat usage, offline behavior, and data size
- Agree on the selected approach with the Ona product and machine-learning teams
- Build the selected approach into the Ona Android application
- Develop any required server-side component
- Ensure capture is linked automatically to the active plot and its metadata
- Support at least one field validation round with an Ona breeding team
- Deliver an inception note covering the current flow, previous attempts, test plan, and device list
- Produce an options report with working prototypes, test results, and a recommended approach
- Deliver a beta build ready for field testing
- Prepare a field validation report covering capture time, success rate, and image quality for plant-stand counting
- Merge the final release into the Ona codebase and provide technical documentation, tests, and a handover session
- Work with the Ona Android developers and machine-learning team through weekly progress calls, code reviews, and pull requests
Requirements
- Native Android development using Kotlin; Java is an advantage
- CameraX or Camera2
- OpenCV and image processing
- Image stitching, panorama generation, feature matching, homography, or image registration
- Gyroscope and accelerometer integration
- Performance optimization for low- and mid-range Android devices
- Offline-first application development and background synchronization
- Server-side image processing using Python, OpenCV, or cloud functions is desirable
- Video frame extraction is desirable
- On-device machine learning using TensorFlow Lite, ML Kit, or ONNX is desirable
- Object tracking is desirable
- Android NDK and C++ are desirable
- Experience in agriculture, field data collection, or applications used in Sub-Saharan Africa is desirable
- Experience with open-source or publicly funded projects is desirable
- Strong written and spoken English
Document Requirements
Application Instructions
- State Ona in-app plot imaging consultancy as the position applied for.
- Submit your CV (with Android camera, stitching and computer-vision work), portfolio links (apps, repositories or demo videos), technical proposal and financial proposal (daily rate and total fee for about 60 working days) as one document named with your last name and first name.
- Email applications will not be considered.
