Dog images recognition and retrieval

Name: Ridnarong Promya

LCSH: Kasetsart University -- Theses. M.Eng. (Information and Communication Technology for Embedded Systems) 2021

Classification :.LCCS: TK7882.P3

LCSH: Kasetsart University. -- Department of Electrical Engineering -- Theses

LCSH: Pattern recognition systems

LCSH: Optical pattern recognition

LCSH: Computer vision

LCSH: Image processing -- Digital techniques

Abstract: Free-roaming dog population survey is very important in the veterinary research area. This thesis investigates two networks, which are Facenet and Dlib for dog identification. These two networks predict embedded vectors which are a reduced dimension vector of images containing distance information that is aligned with dog identities. The full body and face images of dogs are fed into the networks. Our experiments show that facial images provide more discriminative embedded vectors than full body images. The average accuracy of both networks with face and full body images are 91.43% and 87.32%, respectively. Overall, both networks provide good accuracy and promising results for dog recognition applications. This study further do search experiments on various image conditions and search schemes. The scale of image greatly affects performance, and the same image data source is obviously the best search coverage result. Since this study use a pre-trained model with human faces, then transfer those weights and then train with our dog images to improve performance. Head training where only some of the top 2D Convolution layers are trainable and the rest are freeze, provide the best result of improving existing pertain network by the body images average accuracy for search performance is 98.57%. Full training and head transfer scheme model can successfully used in dog image retrieval.

Kasetsart University. Office of the University Library

Address: Bangkok

Email: tdckulib@ku.ac.th

Name: Somying Thainimit

Role: Thesis Advisor

Name: Chalermpol Charnsripinyo

Role: Thesis CO-Advisor

Name: Yasuharu Koike

Role: Thesis CO-Advisor

Created: 2021

Modified: 2025-07-10

Issued: 2025-07-10

วิทยานิพนธ์/Thesis

application/pdf

URL: https://www.lib.ku.ac.th/KUthesis/2564/ridnarong-pro-all.pdf

CallNumber: TK7882.P3 .R53

eng

DegreeName: Master of Engineering

Level: Master's Degree

Descipline: Information and Communication Technology for Embedded System

Grantor: Kasetsart University

©copyrights Kasetsart University

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	ridnarong-pro-all.pdf	2.67 MB

ใช้เวลา

0.020671 วินาที