Cloud Infinite provides image indexing and splitting function through the imageMogr2 API. It can split images into specified sizes and extract the specific block. This feature supports the following processing methods:
Processing during download
Processing during upload
Processing in the cloud
Authorization Guide
When used with a sub-account, corresponding permissions must be granted. For details, please refer to Cloud Object Storage. For other CI permissions, please refer to Cloud Infinite.
Processing during download: Set action to cos:GetObject in the authorization policy.
Processing upon upload: Set action to cos:PutObject in the authorization policy.
Cloud data processing: Set action to cos:PutObject in the authorization policy.
Note：
The resource corresponding to GetObject is the source file, and the resource corresponding to PutObject is the transferred file.
Service Activation
Using this feature will activate Cloud Infinite for you by default and bind a bucket. No role authorization is required, it can be used directly.
Use Limits
When using this API, please first confirm the relevant restrictions. For details, see Specifications and Limits.
Fee Description
This API is a paid service. The incurred fees will be collected by Cloud Infinite. For detailed billing information, see Image Processing Fees.
If processing during download is adopted and the download is performed via the public network, the API call will incur Traffic Fees.
API Example
1. Processing during download
GET /<ObjectKey>?imageMogr2/indexcrop/width/<width>
Authorization: A request header that carries authentication information and is used to verify the legitimacy of requests. The Pic-Operations header must be included in the signature information. For details, see the Request Signature documentation.
Pic-Operations is a JSON format string. For specific parameter details, refer to Persistent Image Processing.
When fileid is the same as ObjectKey, the processed file will overwrite the original file storage; if they are different, the original file and the processed file will be stored separately.
The value of fileid needs to be passed through URL-safe Base64 encoding.
Processing Parameter Description
Parameter
Description
Type
Required
ObjectKey
Object filename, for example folder/sample.jpg.
String
Yes
indexcrop/width/<width>
Specify the length of each area divided by horizontal cuts, in pixels. The value must be a positive integer greater than 0 and not exceeding the image width.
Note: You may only use either the <width> or <height> parameter.
Int
Yes
indexcrop/height/<height>
Specify the length of each area divided by vertical cuts, in pixels. The value must be a positive integer greater than 0 and not exceeding the image height.
Note: You may only use either the <width> or <height> parameter.
Int
Yes
indexcrop/index/<index>
Index value of the image area, default value is 0, which means taking the first block.
Int
No
/ignore-error/1
When processing requests carry this parameter, if operations fail due to oversized files or parameter exceeds limit, it will directly return the original image without reporting an error.
String
No
Note:
If the specified index value is greater than the number of regions formed after slicing, the original image will be returned.
When both <width> and <height> are specified with valid values, the <height> parameter takes precedence.
Practical Example
Note:
The examples in this documentation contain only processing during download. This type of processing will not save the processed images to the storage bucket. If you have a need to save, you can check the Image Processing Mechanism document and configure process during upload or processing in the cloud.
Case 1: Index Slicing in Horizontal Direction
The original image URL of the example is as follows:
Case 2: Extract a Square Area by Combining Horizontal and Vertical Index Slicing
Slice the image horizontally at 100 pixels and connect using the Pipeline Operator, then slice vertically at 100 pixels and take the first segmented area. The processed image link is as follows:
