Product Introduction
Abnormal image detection helps you determine whether images stored in COS are impersonating files (e.g., containing hidden non-image content like videos), to prevent your COS from being maliciously exploited as a video source resulting in traffic theft.
Note:
When using abnormal image detection, please first confirm the relevant limitations and Region. For details, see Usage Limitations.
Prerequisites
You have activated the COS service, created a Bucket, and uploaded images to the Bucket. For detailed operations, see Bucket files. You have activated the Cloud Infinite service and bound a Bucket. For specific operations, see Bucket Binding.
Usage
You can detect abnormal images in the following ways:
Automatic Detection During Upload
You can use workflows to automatically detect abnormal images. By enabling workflows for a specific Bucket or path, files uploaded to that Bucket or path will automatically undergo abnormal image detection. It supports redirecting detected abnormal images to a specified location or provides the option to automatically delete them. For details, see Workflow Documentation. Call the API for Detection
You can call the Abnormal Image Detection API to detect images stored in COS. For usage details, see the API documentation. Perform Detection on Batch Images in COS
You can create a batch processing task to perform one-time batch detection on images in a specific Bucket or path. It supports redirecting detected abnormal images to a specified location or automatically deleting them. For details, see Batch Processing Task Documentation.