Method for calculating loss function of fallen leaf detection prediction frame

A technology of loss function and detection algorithm, which is applied in computing, computer parts, instruments, etc., can solve the problems of only focusing on the loss of a single leaf, not considering the complex situation of leaf falling, and the unsatisfactory results of detecting leaf falling.

Pending Publication Date: 2021-08-17
CHINA UNIV OF MINING & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The existing target detection algorithm can complete the detection task of a single fallen leaf very well, but the result of detecting fallen leaves is not ideal. The main reason is that the prediction frame loss function of the existing algorithm only focuses on the calculation of the loss of a single fallen leaf. Did not take into account the complex situation of fallen leaves

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
  • Method for calculating loss function of fallen leaf detection prediction frame
  • Method for calculating loss function of fallen leaf detection prediction frame
  • Method for calculating loss function of fallen leaf detection prediction frame

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0033] The present invention proposes a new prediction frame loss function—AIoULoss (gravitational loss function) to solve the problem of detecting fallen leaves, and the present invention will be further described below.

[0034] The present invention can be realized through the following technical solutions:

[0035] A method for calculating a prediction frame loss function suitable for leaf fall detection, comprising the following steps:

[0036] Step 1: First calculate the Euclidean distance between the center points of the target frame of the calibrated data set, and calculate its average value l Avg , and let l Avg = δ. δ represents a decision threshold.

[0037] Step 2: Assuming that there are N prediction frames in total, the distance between the N prediction frames will be calculated below. Introduce parameter l Pre Represents the Euclidean distance between the center points of the prediction frame, and sorts the calculated distances to get:

[0038]

[0039]...

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 provides a calculation method of a prediction frame loss function-gravitation loss function (AIoU Loss) for fallen leaf detection, aiming at the problems that an intelligent cleaning device is difficult to detect flakes of fallen leaves and low in efficiency. Distance parameters between prediction frames are introduced, and the minimum external rectangular frame is used for replacing the original prediction frame meeting the condition, so that the effect of predicting the fallen leaves in the flakes is achieved. According to the intelligent fallen leaf sweeping device, the problem of distinguishing single fallen leaves from single fallen leaves during fallen leaf detection is solved, the intelligent fallen leaf sweeping device can more accurately position fallen leaf sites, and the sweeping efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of target detection, in particular to a method for calculating a loss function of a falling leaf detection prediction frame. Background technique [0002] At present, with the development of computer vision technology, object detection has become a current research hotspot, and has broad application prospects in the field of sanitation. The cleaning of fallen leaves on the road surface is a difficult point in sanitation work, especially the fallen leaves in parks and narrow roads can only be cleaned manually, which is time-consuming and laborious. Therefore, using target detection technology to realize intelligent cleaning of fallen leaves on the road surface has become a new way. [0003] The needs of falling leaf detection are different from general object detection. The distribution of single fallen leaves and fallen leaves is uneven. For the cleaning of fallen leaves, it is only necessary to detect the ...

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
IPC IPC(8): G06K9/00G06K9/62
CPCG06V20/41G06V20/52G06F18/22G06F18/25
Inventor 缪燕子张宗伟王贺升赵忠祥代伟王啸林翟煜史延诺王志铭王玥
Owner CHINA UNIV OF MINING & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap