Unlock instant, AI-driven research and patent intelligence for your innovation.

Image Salient Object Detection Algorithm Based on Object Prior

A target detection algorithm and salient technology, applied in computing, computer parts, instruments, etc., can solve the problems of discontinuity of salient regions, ignore the integrity of salient targets, etc., achieve high accuracy and recall rate, and maintain the effect of smoothness

Active Publication Date: 2021-08-03
TIANJIN UNIV
View PDF11 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

These existing methods have achieved good saliency detection results, but these methods often calculate saliency values ​​for each small area, ignoring the integrity of salient objects, so there is an internal discontinuity problem in the detected saliency area

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image Salient Object Detection Algorithm Based on Object Prior
  • Image Salient Object Detection Algorithm Based on Object Prior
  • Image Salient Object Detection Algorithm Based on Object Prior

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] Embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0038] Such as Figure 1 As shown, a kind of object prior-based image saliency target detection algorithm flow diagram of the present invention, its specific details are as follows:

[0039] Step 1. Using SLIC superpixel segmentation algorithm [4] Segment the image into N superpixels {R i}, and then calculate each superpixel R according to the spatially weighted region contrast i The initial saliency value S i initial ,Calculated as follows

[0040]

[0041] Among them, ci and c j Respectively represent the superpixel R i and R j Values ​​on the CIE-Lab color space, p i and p j Respectively represent the superpixel R i and R j The normalized spatial position value of , σ p Represents a constant controlling the global contrast weight;

[0042] An initial saliency map is thus obtained;

[0043] Step 2. Using the prior knowledge ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an image saliency target detection algorithm based on object prior. In step (1), the image is divided into N superpixels, and then the initial saliency value of each region is calculated according to the spatially weighted region contrast, thereby obtaining An initial saliency map; step (2), a single input original image generates multiple target candidate blocks through an algorithm, and screens out a series of high-quality target candidate blocks; step (3), by comparing each target candidate block with step (1) Obtain the overlapping rate of the initial saliency map, and calculate the score of each target candidate block covering the salient target; step (4), take the score of each target candidate block as the weight, and weight the filtered target candidate blocks to obtain The saliency map S at the target level obj ; Step (5), obtain the final saliency value S by solving the minimized energy equation. The present invention can simultaneously maintain high accuracy rate and recall rate in different data sets, and can accurately locate salient objects.

Description

technical field [0001] The invention relates to the technical field of digital image processing, in particular to a method for detecting prominent objects in an image. Background technique [0002] With the development of information technology and the increasing popularity of intelligent terminal products, hundreds of millions of multimedia information data are continuously generated and disseminated every day, which brings great challenges to image and video processing. In the face of massive information in the era of big data, how to effectively improve the efficiency of computer image analysis and processing has become a hot spot for researchers in the field of computer vision. [0003] Neuropsychological studies have found that the human visual system often screens out the most interesting areas first when processing complex scenes, and prioritizes these areas, so as to achieve rapid analysis and understanding of complex scenes. Inspired by this mechanism, researchers ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/46G06K9/44G06K9/62
CPCG06V10/34G06V10/462G06F18/22
Inventor 周圆毛爱玲霍树伟张天昊李绰
Owner TIANJIN UNIV