About Me

I am currently a third-year master's student at Beijing Forestry University. My mentor is Prof. Kan Jiangming from BFU-Key Laboratory of State Forestry Administration on Forestry Equipment and Automation Labs. Prior to that, I was an undergraduate student in the College of Science at China University of Mining and Technology (Beijing), majoring in Information and Computing Science and minoring in Computer Science.

My main focus lies in Computer Vision, especially in environment perception in complex forest spaces. I try to optimize the pipeline of 3D representations of forests generated based on images and extract structural information of trees in that representation for automated forest inventory. My research work also involves the development of automated forestry and agricultural equipment using RGB or RGB-D as sensors.

Throughout my academic journey, I've consistently held myself to a high standard of responsibility and purpose. My academic performance has consistently ranked at the top within my college, boasting a GPA of 92.8 and securing the 1st position in my school.

My approach to learning extends beyond the classroom. I actively engage in specialized knowledge within my field while also immersing myself in the latest research papers to broaden my horizons. Additionally, I seek opportunities outside of school to learn from high-performing teams, using their expertise to address my own shortcomings.

My dedication to learning is reflected in my academic achievements, and I firmly adhere to the principle that hard work compensates for any deficits. In the rigorous academic environment, I maintain a strong sense of academic integrity, striving for excellence and perfection in every aspect of my work. I relentlessly pursue higher standards in my academic research endeavors.

Publications


“Location of pruning points in tall spindle apple trees with RGB-D based on SOLOv2.”
Dr. Siyuan Tong; Jiaming Zhang; Prof. Feng Kang*; et al.
“Defocus blur detection based on Transformer and complementary residual learning.”
Shuyao Chai; Prof. Xixuan Zhao; Jiaming Zhang; Prof. Jiangming Kan*; et al.
“A Lightweight Forest Scene Image Dehazing Network Based on Joint Image Priors.”
Prof. Xixuan Zhao; Yu Miao; Zihui Jin; Jiaming Zhang; Prof. Jiangming Kan*; et al.
“Tea Bud and Picking Point Detection Based on Deep Learning”
Junquan Meng; Prof. Yaxiong Wang; Jiaming Zhang; et al.
“Impact of Weather Insurance on Household Production and Savings”
Dr. Xiaoling Zhang; Jiaming Zhang.
“Numerical simulation based on the CIR model and positive comparative analysis of set-up participation”
Shilei Gao; Anqi Li; Jialu He; Heng Yuan; Jiaming Zhang.

Work In Progress


“An Advanced Software Platform and Algorithmic Framework for Mobile DBH Data Acquisition”
Jiaming Zhang; Dr. Siyuan Tong; Prof. Jiangming Kan*; et al.
“A Novel Forestry Information-collecting Mobile System.”
Prof. Shun Yao; Jiaming Zhang; Prof. Jiangming Kan*; et al.

Talks


Efficient forestry information-collecting system based on multi-modal mobile robot and its application
Invited speaker @ the 8th China Forestry Academic Conference, 08/2023

Awards


Research Experience

Backend Server Development for Multi-user Platform

  • Designed and implemented a robust server backend using I/O multiplexing and thread pooling models to support multiple user logins simultaneously;
  • Developed a socket-based networking module with features like "Broken-point Continuingly-transferring" and incorporated Reactor architecture for efficient request handling;
  • Integrated MySQL database for user data storage and retrieval, ensuring data integrity and security;
  • Led the development of a command-line interface for system components, enhancing user interaction and system control;
  • Collaborated closely with task delegates for system co-tuning, ensuring optimal performance and timely business maintenance.

Dunhuang Gobi Wind-Sand Jumping Study Based on 3D Gravel Morphology

  • Engineered a bespoke photogrammetric apparatus specifically for intricate three-dimensional analysis of gravelly terrains, facilitating accurate digital modeling and extraction of their morphological characteristics;
  • Introduced a point cloud segmentation methodology tailored for gravel, enabling the extraction of gravel point clouds across varied scales from irregular terrains and densely packed scenarios;
  • Conducted comprehensive analyses on the high-resolution three-dimensional model of gravel surfaces, including shape, roundness, and sphericity, leading to the determination of the gravel's fractal dimension and computation of the Hurst coefficient.

Intelligent Defect Detection Based on Computer Vision

  • Implemented an image preprocessing algorithm to filter out stray light, enhancing the detection rate of the algorithm;
  • Optimized imaging quality using the pinhole camera model, improving image performance metrics such as NIQE and information entropy.

Deep Learning & RGB-D in Apple Tree Pruning

  • Devised a robust algorithm, integrating deep learning with RGB-D, to pinpoint pruning locations on dormant tall spindle apple trees using Cascade Mask R-CNN SW-T;
  • Developed an image post-processing method to determine junction points, potential pruning points, and branch diameter endpoints;
  • Designed an optimized pruning strategy for robotic arms based on actual pruning rules and robotic arm characteristics.

Automatic Measurement of Tree Diameter at Breast Height based on Computer Vision

  • Developed and validated an efficient method for measuring tree diameter at breast height (DBH) using Structure from Motion (SfM) photogrammetry;
  • Implemented a comprehensive software suite for SfM-based reconstruction and a multi-modal robotic platform for forest resource data acquisition;
  • Instituted a real-time forest metric capture system using an RGB-D camera and wireless transmission.

Teaching Experience

Machine Vision(For MS / PhD)

Digital Electronics (B.Eng)

Community Service

The Chinese Mental Wellness and Suicide Prevention Society

This is a non-profit organisation registered in New York, dedicated to the development, application and promotion of multimedia tools for self-management of mental health. We have established a humanistic mental health care network system, including a cloud-based platform and a community assistance platform, to support and facilitate interactions with other research and education public interest platforms, organisations and enterprises serving the community.

Community Support Volunteer during COVID-19 Blockdown Times

During the peak of a major outbreak in my hometown, I dedicated myself as a Community Support Volunteer. I actively contributed to resource mobilization efforts by transporting and distributing food, assisting elderly residents in need, and ensuring the timely delivery of vital medications. In this challenging environment, I seized the opportunity to make a difference by empowering individuals and fostering community solidarity. This experience not only showcased the resilience and determination of the human spirit but also reinforced my belief in the power of collective support during times of crisis.