Contact person: Business Client

Phone:Show

Email:Show

Location: Singapore, Singapore

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I want a simple web application that opens the phone camera or uploads an image from folders, lets the user snap a soil sample in-app (JPEG only), and instantly returns the closest Munsell colour code. No machine-learning or AI shortcuts—just a direct comparison against a built-in digital scan of the official Munsell Soil Color Charts.

Core flow
• Camera capture → app stores the photo as JPEG.
• reads the pixel data, compares it against the chart values, then displays the exact hue, value, and chroma code (e.g., 10YR 5/3).
• User can save or copy the result for field notes.

Key requirements
• Colour matching logic must rely only on the chart dataset; please avoid TensorFlow, CoreML, or any AI libraries.
• Ambient-light correction and a white-balance prompt are welcome so results stay consistent in outdoor conditions.
• All processing happens on-device; no server calls needed.

Deliverables
- working web application.
- Source code with clear comments explaining the comparison algorithm
- A JSON or similar file containing the digitised Munsell reference colours
- Quick start guide describing build steps and how to replace or update the chart data

Acceptance criteria
The app should identify sample swatches from the printed Munsell book under daylight within one step of human visual selection at least 90 % of the time." (client-provided description)