
Elphel Inc Profile last edited on: 12/18/2020
CAGE: 4Y6R2
UEI: LNZSLVEJHBF9
Business Identifier: High-performance digital cameras, image processing, 3D imaging and machine learning. Is this YOUR Company?
Ensure accuracy and completeness of YOUR Company Profile by completing the brief Survey Instrument attached
Do you know about this Awardees?
Let us encourage you to provide any data which would enhance the completeness of this firm's profile.
Location Information
Location: Single
Congr. District: 02
County: Salt Lake
Congr. District: 02
County: Salt Lake
Public Profile
Elphel develops high-performance digital cameras, image processing, 3D imaging and machine learning. The firm offers free software and open hardware. Elphel imaging systems are primarily used for scientific applications that require designs to be user-modifiable at all levels - from the hardware and FPGA code to the system and application software. Elphel cameras offer a combination of high performance and complete openness at all levels from mechanical and circuit design, FPGA RTL code up to all levels of the system and application software. All the design files are publicly available at Elphel web sites and through the popular GitHub repository. These features make Elphel products attractive for scientific research and innovative products. Elphel technology has been referenced in over a hundred scientific publications, and at least six US patent applications that use or reference Elphel products.
Since 2012 Elphel has been developing methods of precise camera calibration and designing photogrammetric multiple-view cameras with thermally compensated SFE resulting in 0.05 pix reprojection error for 5 MPix sensors. We developed modified checkerboard pattern where each edge consists of 2 arcs - this modification from traditional straight lines checkerboard makes power spectrum uniform and improves accuracy of the indirect measurement of the PSF for subsequent optical aberration correction of the acquired images.
From 2016 till present we are working on very long range 3D reconstruction and achieved 0.05 pix disparity resolution by combining several technology components developed by Elphel.
Use of multiple image sensors (4+) instead of conventional binocular stereo.
Advanced calibration, frequency-domain rectification, aberration correction, filtering, and 2D phase correlation.
Training and inference of the Deep Neural Network to predict the depth map from 2D correlation results. DNN analyzes context, extracts object edges, separates background, fuses depth map with textures.
Elphel MNC393-XCAM camera is a narrow baseline 3D perception system providing 10% ranging accuracy at 2000 m, by utilizing the achieved 0.05 pix disparity resolution.
Extent of SBIR involvement
Synopsis: Awardee Business Condition
Employee Range
1-4Revenue Range
0VC funded?
NoPublic/Private
Privately HeldStock Info
----IP Holdings
1-4Awards Distribution by Agency
Most Recent SBIR Projects
Year | Phase | Agency | Total Amount | |
---|---|---|---|---|
2021 | 1 | NASA | $124,972 | |
Project Title: Long-Range Thermal 3D Perception in Low Contrast Environments | ||||
2021 | 2 | AF | $793,487 | |
Project Title: Thermal 3D Perception for Autonomous Navigation | ||||
2019 | 1 | AF | $109,996 | |
Project Title: Helmet Mounted LWIR 3D Perception System with Machine Learning |
Key People / Management
Andrey Filippov
Company News
There are no news available.