allbeingsfuture/skills@image-understanding

To extract text from an image using OCR, you can use this Skill to send photos, screenshots, or scanned documents to vision models. It processes the image and returns extracted text while preserving the original format where possible.

Yes, you can analyze a product photo and answer questions about its features using the visual Q&A mode. This mode allows you to ask targeted questions about image content, enabling detailed product feature extraction and chart interpretation.

You can convert a whiteboard photo into editable notes by applying the OCR and image description modes. This Skill analyzes whiteboard captures to extract all visible text and generate scene-level narratives for meeting documentation.

Yes, you need a Dashscope (Aliyun) API key to perform visual question answering. This Skill requires the API key to access and utilize the qwen-vl-plus and qwen-vl-max vision models for processing image inputs.

The best way to generate detailed descriptions of scene images is using the describe mode within this Skill. It analyzes images to produce scene-level narratives covering people, objects, colors, and overall composition.

Image understanding works with both local image files and URLs. This Skill supports local file base64 embedding as well as direct image URL input for flexible integration into your visual analysis workflows.