Hyeongju Lee 이형주
KRJPEN CV & portfolio
Research Master's program in Creative Informatics, Kyushu Institute of Technology · advisor: Prof. Min-Chul Lee

Reconstructing lens array images in 3D with geometric optics, without training

ICMV 2026 Best Poster Award4 first-author papers at international conferences
01 Try it

Capture with a lens array, rearrange, reconstruct

Your browser captures a 3D scene through a lens array and computes it with the paper's method (RIM). Click the screen to start, with sound.

Objects
Lens array
Sensor
Reconstruction plane

FigureThe paper's method, RIM, ICMV 2026 Best Poster Award · How it was built →

In step 3 you can move the reconstruction distance yourself

–
02 Research flow

From input to result metrics

The three lines of papers, drawn side by side in the same stages. Click a node to see the method and results at that stage.

My method (first author)Input data
  1. Data
    Evaluated on 7 Stanford Lytro light field scenes. The elemental image each lens forms is only a few pixels wide.
  2. Method
    A one-to-one rearrangement that swaps (lens index, pixel position under the lens) (Ray-based Inverse Mapping). Under telecentric capture and orthographic recapture there is no interpolation, so the quantization error is zero.
    Presented
    ICMV 2026, Budapest · Best Poster Award
    More
    RIM and iRIM page
  3. Method
    The rearranged views are used as the first stage of computational integral imaging reconstruction (CIIR) and compared with using the elemental images as they are.
  4. Result
    On the 7 Lytro scenes: PSNR 23.31 → 25.70 dB, SSIM 0.741 → 0.877, and pretrained YOLO detection mAP50 0.665 → 0.776.
  1. Data
    9×9 elemental images, the camera parameters and the depth to reconstruct.
  2. Method
    The search that reconstructed again and again to find the origin with the least quantization error became a calculation from the camera parameters alone. In the paper's experiments it gave the same origin and SSIM as the iterative search.
    Presented
    ITC-CSCC 2025 (Seoul, poster) · ITC-CSCC 2026 (Bangkok, oral)
  3. Result
    For 9×9 elemental images, the reconstructions needed to choose the origin dropped from 81 to 1.
    Stack
    • Python
    • NumPy
    • CuPy
  4. Result
    CPU 0.31 s (measured in the paper), GPU (CuPy, RTX 3090) 0.0057 s. The 25.45 s of the iterative search is an estimate, 0.3142 s per run × 81.
    Code
    origin-point-optimization-vcr
  1. Data
    Views rendered from several positions.
  2. Method
    Because RIM is one to one, it also runs backwards: iRIM synthesizes the elemental image of the virtual lens array you want from multi-view images.
    Presented
    ICCAS 2026, Sapporo · poster
  3. Result
    A 6×6 configuration matched to a 24-inch FHD panel (320×180 lenses). Every view has 100% effective coverage, against 6.7% for the corner views of a simulated lens array camera.
ImprovedReconstructions needed to pick the origin: 81 → 10.31 s measured in the paper · the iterative 25.45 s is an estimate
03 Topics

Four topics

The focus sweeps through depths, then each object is shown with its box and depth
Master's thesis (in preparation)

Real-time integral imaging: detect once, refocus only inside each box

From a single real Lytro lens array image, the GPU estimates a box and a depth for each object in real time. Instead of running detection at every depth plane, it detects once and refocuses only inside each box.

Lytro data: Stanford Lytro Light Field Archive, occlusions_25 (Raj et al., 2016)

16.4 ms per real Lytro frame (61 fps, RTX 3090)
Input and result for a dark photo, shown in turn
Patent application in preparation

Image enhancement for low-light and hazy scenes

Photos taken in the dark or in haze are processed one at a time and made clear. Above: a dark photo from public data.

Public data: SI-HDR (Hanji et al., SIGGRAPH 2022, CC BY 4.0)

One image at a time
Elemental image of a 6 × 6 lens array synthesized with iRIM
ICMV 2026 Best Poster Award · ICCAS 2026 poster

RIM and iRIM: turning angular samples into spatial resolution

RIM rearranges the pixels of a lens array image one to one, without interpolation, into views; iRIM runs it backwards to synthesize the elemental image of a virtual lens array from multi-view images. Step 2 of the lab above is RIM.

PSNR +2.39 dB · SSIM +0.14 on 7 Lytro scenes (as reported in the paper)
Resolution chart reconstructed with a fixed originResolution chart reconstructed with the optimized origin
ITC-CSCC 2025 poster · ITC-CSCC 2026 oral · code public

Origin point optimization: sharper, then faster volumetric reconstruction

Optimizing the origin point reduces the quantization error of volumetric computational reconstruction (VCR), and the origin is then computed from the camera parameters without an iterative search.

81 reconstructions → 1 (paper: 0.31 s; the iterative search an estimated 25.45 s)
04 Papers

4 first-author international conference papers, and a domestic presentation

  1. ICMV 202619th · Budapest

    Angular-to-Spatial Information Redistribution via Ray-based Inverse Mapping

    Hyeongju Lee, Myungjin Cho, Min-Chul Lee · poster Best Poster Award · RIM
  2. ICCAS 2026Sapporo

    Virtual Lens Array Synthesis from Multi-View Images via iRIM

    Hyeongju Lee, Myungjin Cho, Min-Chul Lee · poster · iRIM
  3. ITC-CSCC 2026Bangkok

    Accelerated Origin Point Optimization for Volumetric Computational Reconstruction

    Hyeongju Lee, Myungjin Cho, Min-Chul Lee · oral · code
  4. ITC-CSCC 2025Seoul

    Image Enhancement of Volumetric Computational Reconstruction Using Origin Point Optimization

    Hyeongju Lee, Gilsu Yeo, Myungjin Cho, Min-Chul Lee · poster · code
  5. OSK 2024Korea · Photonics Conference

    원점 최적화를 이용한 체적 컴퓨터 재생의 이미지 향상 (image enhancement of volumetric computational reconstruction using origin point optimization)

    Hyeongju Lee and 7 co-authors, Optical Society of Korea · domestic poster (the groundwork for the ITC-CSCC 2025 paper)
05 Research path

From a course project to master's research

  1. 2022
    Starting point in vision and MLDepth estimation with correlation filters and VCR, a course project in MATLAB (code), is where this research started. Alongside it: an MNIST CNN in MATLAB with backpropagation and momentum gradient descent written by hand, no deep-learning framework. My interest in vision began with my graduation project, Smart Mirror (team of two: Raspberry Pi, OpenCV LBPH face recognition, Kakao speech API).
  2. Aug 2024 – Mar 2025
    Undergraduate researcher, 3D Image Processing Systems Lab, Hankyong National UniversityAdvised by Prof. Myungjin Cho. A first-author poster at Photonics Conference 2024, the groundwork for the ITC-CSCC 2025 paper
  3. Apr 2025 – now
    Master's program, Kyushu Institute of TechnologyLight field and integral imaging 3D reconstruction, advised by Prof. Min-Chul LeeFinishing my master's in March 2027
  4. 2026
    Rotary Yoneyama Memorial Foundation ScholarSelected through a Japanese-language interview in Rotary District 2700
Contact

For roles and collaboration, please write by email.

I am looking for roles in data engineering, backend development and vision data pipelines.

Finishing my master's in March 2027