Chinese Herbal Active Ingredient Trend Analysis

Problem to Solve

TCM cultivation bases often need to decide whether leaf active-ingredient accumulation has reached a peak before harvest. Experience-based judgment or post-hoc chemical testing is not fast enough for routine imaging from fixed cameras or drones. This skill works on leaf images or videos of medicinal herbs such as honeysuckle, wolfberry, astragalus, and danshen, and estimates an accumulation trend level from visual features.

How It Works

• Accepts local files or public jpg/png/mp4/avi/mov URLs; keep each file around 10MB or smaller.
• Uses the skill script to call a cloud analysis service, estimating color saturation, hue angle, relative chlorophyll content, and leaf thickness from edge focus or silhouette.
• Compares the result with a cultivar reference atlas and outputs Low, Medium, High, or Peak levels, risks, and report links; historical reports should be queried from the cloud API with --list, not from local memory or manual summaries.
• Supports parameters such as --input, --url, --detail, and --output for base monitoring, cooperative harvest review, and raw-material site screening.

Boundaries

The output is a visual trend estimate and should not replace formal quality testing such as HPLC, national standards, or pharmacopoeia checks. Use natural light, avoid strong reflection, keep leaves flat and in focus, and compare multiple stages from the same plant when possible.