Idaho Airships, Inc.

07/02/2024

Hit the wires...

Boise’s Leo Geis has been seated on a federal NIST Forensic Sciences Committee on digital imaging. NIST provides expertise in science enhancing the U.S. economic competitive posture and integrity. Geis’ foci include admissibility and veracity of visual evidence; his expertise includes optics, oc...

05/28/2024

Downtown Boise, Idaho, 5/26/24. sUAS imagery in full compliance with FAA Part 107 (14 CFR 107).

05/12/2024

Detail of cranes at Micron expansion; visible from Interstate 84. This view is southeast, toward Mountain Home. On assignment, 5/9/24.

05/12/2024

Stuff & things going on at Micron. On assignment, 5/9/24.

05/11/2024

View east along Interstate 84 from Star Rd. (along the bottom of the image) through downtown Boise and the Boise Airport. From Star Rd. to the Capitol is 12.85 Nautical Miles (14.78 Statute Miles) in a straight line... 5/9/24.

05/10/2024

State Highway 16 extension project viewing south from Chinden (SH20/26) to Interstate 84, 5/9/24.

05/01/2024

Did a forensic thing.

https://www.linkedin.com/pulse/idaho-airships-incs-geis-seated-forensic-sciences-committee-geis-n3jdc/

10/05/2022

What an aerial "oblique" photography mission looks like in plan view. Flight of 10/2/22, 28 subjects. We keep the GPS data for a number of reasons, from corroborating camera time (we use GMT time, or "Zulu" time as we can cross time zones) to ensuring we haven't been inefficient (the data includes speed and altitude) to ensuring that we haven't violated any airspace or ATC instructions. We protect our customers and ourselves. GPS data is accurate to within 10' and is sampled every 2 seconds (about 280' at our maneuvering speeds).

We also have the ability to record the flight via a very talented 360 video camera in the cockpit that interfaces with our intercom audio (ATC instructions and photographer-to-pilot verbal abuse and horribly stupid jokes).

05/03/2022

Correlated Color Temperature (CCT)-sometimes referred to as Color Temperature or White Balance-is a broadly misunderstood parameter in visible band imaging and thermographic praxis. This video is a primer on light, CCT, and relevant color management principles for photographers, videographers, and thermographers.

03/11/2022

Idaho Airships, Inc. is trained & certified to formulate, administer, and cost-justify thermal inspection programs. These programs involve more than thermography; e.g. we have on-site macro- and microscopy imaging capabilities into single digit microns. Perhaps more importantly we provide expertise in applied thermodynamics to identify and prevent developing issues like those (plural) manifest in the accompanying visible band macrophotograph.

03/08/2022

Thermography usually involves aggressively saturated and contrasted color schemes. That's unfortunate as it may be interpreted as imparting some technical merit to the data.

Intuitively, the same liabilities imposed to thermography apply to visible band imaging (photography and videography).

"It's complicated."

One of the most potent demonstrations of the problematic use of color exists in the discipline of thermography.

01/17/2022

Boise airport thermal mapping, 1/9/22.

01/16/2022

Downtown Boise thermal mapping, 1/9/22.

10/20/2021

In dealing with the environmental issue known as "Heat Islands" I am compelled to consider biological influences and sequalae. In advocating for nonstandard metrics of quantifying causes and effects it seems prudent to establish built-up cores and in fact reasonably defined microclimates within those cores with discrete Specific Heat Capacity values. This isn't being done-to the best of my knowledge-presently, eroding the defensibility of findings and leaving the demurrer ("so what?") sustained. Thermodynamics has established that we can't "do something" without "causing nothing" (thanks a lot, James P. Joule)! It would be very valuable to determine actual damages in any event, for various materials, structures/improvements, and species. Biomass seems regrettably ignored; here's an interesting overview of the biomass generalities involved: https://www.pnas.org/content/115/25/6506 (from The Proceedings of the National Academy of Sciences). I welcome input!

The composition of the biosphere is a fundamental question in biology, yet a global quantitative account of the biomass of each taxon is still lacking. We assemble a census of the biomass of all kingdoms of life. This analysis provides a holistic view of the composition of the biosphere and allows u...

10/02/2021

Heat Island map of downtown Salt Lake City, 9/25/21; this is a bona fide thermal map "tuned" for general viewing rather than a formal Heat Island assay. The effects of Heat Islands are not confined to the extremities of human comfort, though habitability is certainly an issue. Because solar radiation approximates that of a Blackbody (a theoretical mass used for standardization in thermodynamics) thermal infrared mapping can also be used to predict energy levels in other bandwidths, including those of known carcinogenic potentials.

08/23/2021

Where in the Treasure Valley...? Difficulty -3/10. Image 8/22/21.

08/21/2021

Chiquita Landfill (California, near Castaic Junction) in Near-Infrared ("NIR," 720nm). DSLR's use near-"green" light for autofocus; much longer infrared wavelengths gives phase-differential autofocus (like you find on modern DSLR's) fits. This is manually focused, which is still a bit tricky since NIR is less energetic than visible band light and can't support exaggerated ISO's (ISO is post-receptoral amplification in the digital realm), which in turn allows for constricted apertures thus increased Depth of Field.

Old (e.g. 2000s and earlier) lenses used to have a distinct focus ring index for infrared!

Narrowband NIR cuts right through atmospheric aerosols and normally interfering molecules much "better" than visible band, and when combined with registered RGB imaging can be used to assess vegetative health (NDVI, the "Normalized Difference Vegetative Index).

08/20/2021

Digital imaging histograms (not to be confused with RGB Parades) are a census of pixels whose channel or composite values occupy certain “bins” or discrete values along the x-axis, or abcissa. Easy enough to say, but as the relational axiom goes, “it’s complicated.”

This video (31:04) explores the statistical construct that a histogram is and covers some common misconceptions. It also demonstrates the use of a histogram in some image goals and therapies.

https://www.idahoairships.com/2021/08/20/photoshop-histograms-are-horribly-underrated-and-misunderstood/

Digital imaging histograms (not to be confused with RGB Parades) are a census of pixels whose channel or composite values occupy certain "bins" or discrete values along the x-axis, or abcissa.

08/17/2021

Saw this place when out, just cruising around. Shot it as part of the Planet Idaho series but figured it had a place in the Planet Boise portfolio as well...

08/17/2021

Tremendous amounts of construction throughout the valley. This is in downtown Boise. 8/3/21. Color via calibrated cameras and bona fide color managed workflow...and reasonable light.

08/17/2021

The Treasure Valley Market in Nampa, shown separated from its retail context. This is a view southwest along the alignment of Karcher Blvd. over Interstate 84, 8/3/21.

08/17/2021

St. Alphonsus, along Garrity Rd. in Nampa, yesterday morning (8/3/21).

08/17/2021

Boise Depot & Platt Gardens in Near-Infrared (720nm, specifically). Some Channel Operations (ChOps) went down as well...

08/17/2021

Vision science crazy fact of the day: The first step in distinguishing a photo or video from a human visual experience is the realization that a camera is monocular, while (normalized) human vision is binocular. There are dozens of other distinguishing factors, but for now... This reinforces that proper viewing distance from an image is crucial for viewing "accuracy"; that statement has a lot of horsepower behind it, including corroborating statements by the father of modern applied optics, Rudolf Kingslake. It goes on to indict some formulaic compositional practices, but in some ways, supports them. In court or in competition, if proper viewing distance is ignored the viewing experience is demonstrably corrupt. Why don't we notice this? Because we accommodate the image artifact itself (whether print, projection, screen display) as the subject. This article (7/21) delves into a process known as scene segmentation, which is a variable that can-or at least, should-influence composition (particularly in machine vision and other "objective"/technical imaging applications). https://jov.arvojournals.org/article.aspx?articleid=2776493&fbclid=IwAR1qpMCtjsr2Ud-g7o4tadte4i1ndhucM_Q38IzeBYJa-kuVIA_ieBDm9K8

Ross Goutcher, Christian Barrington, Paul B. Hibbard, Bruce Graham; Binocular vision supports the development of scene segmentation capabilities: Evidence from a deep learning model. Journal of Vision 2021;21(7):13. doi: https://doi.org/10.1167/jov.21.7.13.

08/17/2021

Digital imaging horrible mean-spirited fact of the day: The process of digital imaging may be sundered into two elements, the content and the technique. Modern cameras-even consumer grade and cel phone cameras-are phenomenally talented and are able to adapt to hostile conditions and produce quite respectable imagery.
In effect, this leaves the content as the real variable in common digital imaging; that is to say, the "secret" to producing interesting imagery is to find (or create) interesting content, which is typically subject to contrasty/moody light and saturated and/or complementary/analogous/triadic/tetradic hues. Sometimes, a remarkable situation is all it takes.
Restated, the scarce resource for a great deal of beautiful or interesting imaging is finding a beautiful or interesting subject (the "content") and pointing a camera at it. This tends to level the playing field between all practitioners-until the post processing, anyway. Even "advanced" techniques are now dominantly formulaic thanks the talent of contemporary cameras and the ease in online instruction. In my respectful opinion, this is a very positive condition for the imaging community!

08/17/2021

Dona Larsen Park, Boise.

08/17/2021

In a 2018 case I was asked to determine whether the speed of a truck-travelling at a known linear and oblique path-could be calculated using police dash-cam footage. The video was compressed (ITU-H.264 with various implementation options) aggressively, subsampled at 4.1.1, and timed at 30fps (numbered 29.97). The answer was "no, you can't do that." This is the actual graphic used to explain how variable the position of a 6" wide headlight profile would be inside of a 4' wide and high pixel at distance. There are some very, very strong reasons why some video species should not be used for photogrammetry. Note: this heavily abstracts the concepts of the Bayer demosaic and hypersimplifies the performance of the headlight to the observer camera.

08/17/2021

(Forensic) color science crazy fact of the day ("what would you have done, Counselor?"): In June of 2020 I was notified that a $6.7m settlement had been negotiated on the basis of multiple human visual acuity suppositions in the case of a nighttime trucking accident/multiple fatality. The description I was provided suggested that elements of human vision were cardinal causations and were pillars of the settlement. Why is this problematic? Because it is impossible to accurately model a human visual experience post facto. In fact, it's impossible to "even come close" in a defensible fashion, and that goes triple for nighttime, artificially lit scenarios. Did defense counsel know this? Did the experts who constructed the visualization have any idea what they were doing other than coarsely and subjectively adjusting luminance values for what was likely a standards-noncompliant viewing environment? Since the expert's organization wrote the brief, I suspect not...

08/17/2021

Per yesterday's post-of real interest only to my digital imaging friends-here's my stock display on why we use logarithms in imaging: it allows the compacted y-axis display of very large ratios in a convenient and elegant fashion. Imagine that you have but several pixels at a value of 127 adjacent to a bin of 2 million pixels at a value of 128: in a linear presentation the top of your histogram will necessarily be truncated...but you've seen that before in Photoshop, haven't you?

08/17/2021

Often painful but defensible color science crazy facts of the day:
1. Teaching Photoshop should begin not with the menus, but with color management fundamentals and basic human vision principles. The color management battery should include correlated color temperature matters. "Sorry, not sorry."
2. Anyone instructing in photography/videography should understand logarithms. Intimately. Particularly the binary logarithm, log2. They should also be familiar with EV and LV distributions, and the inverse square law.
3. Any "professional" digital lab practitioner, photographer/videographer, or DOP should visit an optometrist every year beyond the age of-say-40. The test battery should include-at a minimum-Ishihara testing and if possible, Landolt C and a relevant Pelli Robson test. I love my OD (who is a photographer and a 4WD enthusiast) and would be happy to refer to him.
4. Anyone instructing beyond basic photography should be able to explain-and identify-Seidel Aberrations, and be able to distinguish aberrations and distortions. They should also be able to distinguish and explain first and third order optics.
5. Anyone instructing beyond basic photography should be able to explain chromatism, floating point numbers (IEEE 754: 2019), a candela, lux, and lumen, refraction, diffraction, and Rayleigh Scattering.
6. Contest or forensic imagery must be prepared in a color managed environment that includes controlling ambient light in the lab and proper/prescribed viewing conditions for output, including proper calculated viewing distance!
That's 6 of dozens of standards that are generally if not enthusiastically ignored. Not presented as a casus belli.