Image Elements Removal

Introduction

Image Elements Removal is tailored for e-commerce images, capable of automatically identifying and removing Semi-Transparent Elements, Specific Name, Character, Color-Blocks with Characters from e-commerce images. It allows for the specification of objects to be removed within the images, achieving efficient and precise processing of large volumes of images, thus simplifying the image editing process.

Use Cases

1.Efficient E-commerce Image Processing:

Eliminate irrelevant visual elements to provide a clearer and more focused product display, enhancing the shopping experience for customers.

2.Multi-platform Product Information Synchronization:

When merchants synchronize product listings across different shops or websites where the products will be shown and purchased, they can quickly remove specific watermarks or text to the image standards that meet the customization requirements of the merchants.

3.Marketing Material Preparation:

Merchants can easily remove existing promotional information and quickly update it with the latest campaign details.

Note: Merchants or platforms can batch remove impurities in images to optimize image quality. It is recommended to use it in conjunction with the "Image Elements Detection API" to first filter out images containing text, logos, watermarks, and text-containing color blocks, and then use the "Image Elements removal API" for batch removal. The Image Elements Detection API focuses on "detection," while the Image Elements removal API focuses on "removal." By calling them in sequence, the removal effect can be maximized, reducing "mistaken removals" and "misidentifications."

Key Features

1. Firstly, define the image subject and background, and when specifying objects for removal, you can choose whether the removal scope is the image subject or the background area. It supports both API and editor access methods to meet different types of needs.

   • Image Subject: The core product area within the image.

   • Image Background: The remaining part of the image other than the product subject.

2. Secondly, define the text, logos, watermarks, and text-containing color blocks to be removed. You can confirm the objects to be removed through the following examples for targeted operations.

Samples

Definable Objects for Removal	Before Removal	After Removal
Character	Character involving the image subject area Character in the image background area, not involving the image subject area	The removal effect of Character involving the image subject area The removal effect of Character in the image background area, not involving the image subject area
Specific Name	Specific Name involving the image subject area Specific Name in the image background area, not involving the image subject area	The removal effect of Specific Name involving the image subject area The removal effect of Specific Name in the image background area, not involving the image subject area
Semi-Transparent Elements	Semi-Transparent Elements involving the image subject area Semi-Transparent Elements in the image background area, not involving the image subject area	The removal effect of Semi-Transparent Elements involving the image subject area The removal effect of Semi-Transparent Elements in the image background area, not involving the image subject area
Color-Blocks with Characters	Color-Blocks with Characters involving the image subject area Color-Blocks with Characters in the image background area, not involving the image subject area	The removal effect of Color-Blocks with Characters involving the image subject area The removal effect of Color-Blocks with Characters in the image background area, not involving the image subject area

Access Methods

The product supports two access methods: API and Iframe Editor.

• Recommended Access Method: Editor access for a better experience and generation effect. The Iframe Editor access capabilities are about to be launched. If you have any needs before the launch, please contact us.

Access Type	Supported Features	Suitable Scenarios	Advantages	Limitations
API Interface Access	1) Remove images based on the input removal scope and elements, and obtain the removed images.	Suitable for users with fewer custom requirements and a larger number of image processing needs; also suitable for users with high requirements for automated processing and easy access.	1) Simple and convenient access compared to the editor.	1) Cannot freely adjust the removal area, and the effect is limited.
Iframe Editor	1) With a user-side operation page. 2) Users can manually smudge/frame/circle the area to be removed within the editor.	Suitable for users who need a high degree of customization or have fewer image processing needs; and at the same time, the system being accessed has a corresponding user interface and interaction.	1) Supports manual selection of the removal area. 2) Cannot batch remove images.	1) Compared with API access method, the development cost of access Iframe Editor is higher. 2) Unable to eliminate pictures in batches.

Pricing

To use the API, you are required to choose and purchase an API resource pack from us on a subscription basis.

• Each resource pack is valid for one calendar year upon successful purchase, and enables you to access the API up to the number of requests specified in the pack. No refunds can be provided.

• If you need to purchase more QPS due to business requirements, please contact us via navigation bar or email us ([e-mail]).

• Resource packs cannot be used across different products. For example, if you need to use both product text translation and image translation, you must purchase separate resource packs for each.

The prices are as follows:

Capacity	Price (USD)	Unit Price(USD)	Maximum QPS
1000 images	22	$0.022 /image	10
10,000 images	220	$0.022 /image	10
100,000 images	2200	$0.022 /image	10

Quick Start

1. Sample Request

In the API parameters, you can choose to remove specified elements (watermarks, logos, text, and text-containing color blocks) in the image subject and non-subject areas.

Java

IopClient client = new IopClient(url, appkey, appSecret);
IopRequest request = new IopRequest();
request.setApiName("/ai/image/removal");
request.addApiParameter("image_base64", "Base64");
request.addApiParameter("image_url", "https://ae01.alicdn.com/kf/Sa78257f1d9a34dad8ee494178db12ec8l.jpg");
request.addApiParameter("non_object_remove_elements", "[1,2,3,4]");
request.addApiParameter("object_remove_elements", "[1,2,3,4]");
request.addApiParameter("mask", "474556 160 475356 160 476156 160 476956 160 477756 160 478556 160 479356 160 480156 160 480956 160 481756 160 482556 160 483356 160 484156 160 484956 160 485756 160 486556 160 487356 160 488156 160 488956 160 489756 160 490556 160 491356 160 492156  160");
IopResponse response = client.execute(request);
System.out.println(response.getBody());
Thread.sleep(10);

2. Sample Response

Java

{
  "data": {
    "imageUrl": "https://nhci-aigc.oss-cn-zhangjiakou.aliyuncs.com/ppc-records%2Fimage-remove%2F805fe1d4-8335-4eaa-b4c1-f9952fea21ef.png?OSSAccessKeyId=LTAI5tCv9DpB7gYic1oGsAyv&Expires=4935905309&Signature=eIJOXyUEBoxqb3VtUJF08xFlcgM%3D",
    "usage": 1,
    "width": 800,
    "class": "com.aidc.service.api.client.image.dto.ImageGenResponse",
    "height": 1259
  },
  "requestId": "2140e7df17192333083756998ebc3f",
  "success": true,
  "resCode": 200,
  "resMessage": "success",
  "code": "0",
  "request_id": "2101725817192333083795305",
  "_trace_id_": "2140e7df17192333083756998ebc3f"
}

FAQ

1. When an object is obscured by text, does the removal of the text support the generation of a completely new part of the obscured area?

Currently, it only supports the removal of the area and performs simple color repair. It does not support the removal of complex object areas. The product is continuously being updated and iterated, and the redrawing capability is expected to be launched soon. If you have any needs, Please contact us via navigation bar or Email [e-mail].

1. If the automatic removal effect of the Image Elements Removal API is not ideal, is there a way for users to manually adjust?

Through the Iframe Editor access method, users can manually smudge/frame/circle the area to be removed within the editor for manual adjustment.

1. Can the Image Elements Removal API recognize and preserve copyright information or important logos in the image during the processing?

It depends on the removal elements and scope specified by the user in the request.