Miniaturization, wide-angle, compactness or low-light performance are required in automotive or mobile device. While conventional design techniques are limited, we introduce methods for designing wide-angle lenses using freeform surfaces illustrated by designs showing improved performances.
Data-driven approaches have proven to be very efficient in many vision tasks, and they are now used for optical parameters’ optimization for application-specific camera designs. Methods such as neural networks are used to estimate camera performance indicators related to the point spread function—such as the root mean square (RMS) spot size—from optical parameters. Such procedures help to understand the connection between optical characteristics and push optical design expertize beyond its limits. We investigate these approaches to model the interaction between the distortion of wide-angle designs and their RMS spot size, which is not explained by aberration theory. Specifically, we test off-the-shelf data-driven methods to determine in which conditions we can establish a model that is able to predict the variations of the RMS spot size along the field of view from the distortion function even in the absence of a mathematical model. Although current methods focus on building accurate models often usable for very specific designs—composed of a few elements only, we present a methodology focusing on more complex and realistic wide-angle designs.
The next generation of sUAS (small Unmanned Aircraft Systems) for automated navigation will have to perform in challenging conditions, bad weather, high and low temperature and from dusk-to-dawn. The paper presents experimental results from a new wide-angle vision camera module specially optimized for low-light. We present the optical characteristics of this system as well as experimental results obtained for different sense and avoid functionalities. We also show preliminary results using our camera module images on neural networks for different scene understanding tasks.
As more and more cameras are used for machine perception, the optical design process still relies on key indicators such as point spread function (PSF), modulated transfer unction (MTF) based on aberration minimization. This process has proven efficient for human vision but is not tailored for machine perception. Given a specific computer vision task, it is not always necessary to target the same key performance indicators (KPIs) than when images are visualized by humans. Moreover, this image quality might change during a camera lifespan with the appearance of defocus for example. It is crucial to be able to determine how this kind of degradation can affect a computer vision task. In this work we study the impact of defocus on 2D object identification and show that, for a certain design, it is not impacted by image degradation under a certain threshold. We also demonstrate that this threshold is higher for lower f-number which makes them better design candidates.
Distortion is one major shortcoming of conventional wide-angle lens. The correction of distortion can maximize field of view in a wide-angle lens and benefit natural scenes. In this paper, we propose a compact wide-angle lens utilizing freeform surfaces to address the shortcoming. The design approaches, including the starting point selection, freeform surface conversion scheme, and system optimization, are discussed in detail. Moreover, freeform diagnostic tools are developed aiming to facilitate system optimization and visualize the critical aspects of optical performance. The simulation results are demonstrated and discussed. Analysis results demonstrate that the design strategies proposed in this paper have been shown to be effective in reducing distortion as well as improving optical performance. We envision that the proposed design schemes will have a positive effect on research and application values for wide angle lens system.
Optical design process consists in minimizing aberrations using optimization methods. It relies on key performance indicators (KPIs), such as point spread function (PSF), Modulated transfer function (MTF), or relative illumination (RI) and spot sizes, that depend on lens elements aberrations. Their target values need to be defined -either for human or machine perception- at early stage of the design, which can be complex to do for challenging designs such as extended field of view. We developed an optical and imaging simulation pipeline able to render the effects of complex optical designs and image sensor on an initial aberration-free image. Extracting files from ray tracing software for simulating the PSF and sensor target information, the algorithm accurately renders off-axis aberrations with Zernike polynomials representation combined with noise contribution and relative illumination. The obtained image faithfully represents an optical system performance from the optics to the sensor component and we can then study the impact of additional aberration introduction.
Data driven approaches have proven very efficient in many vision tasks and are now used for optical parameters optimization in application-specific camera design. A neural network is trained to estimate images or image quality indicators from the optical characteristics. The complexity and entanglement of such optical parameters raise new challenges we investigate in the case of wide-angle systems. We highlight them by establishing a data-driven prediction model of the RMS spot size from the distortion using mathematical or AI-based methods.
The new generation of sUAS (small Unmanned Aircraft Systems) aims to extend the range of scenarios in which sense-and-avoid functionality and autonomous operation can be used. Relying on navigation cameras, having a wide field of view can increase the coverage of the drone surroundings, allowing ideal fly path, optimal dynamic route planning and full situational awareness. The first part of this paper will discuss the trade-off space for camera hardware solution to improve vision performance. Severe constraints on size and weight, a situation common to all sUAS components, compete with low-light capabilities and pixel resolution. The second part will explore the benefits and impacts of specific wide-angle lens designs and of wide-angle images rectification (dewarping) on deep-learning methods. We show that distortion can be used to bring more information from the scene and how this extra information can increase the accuracy of learning-based computer vision algorithm. Finally, we present a study that aims at estimating the link between optical design criteria degradation (MTF) and neural network accuracy in the context of wide-angle lens, showing that higher MTF is not always linked to better results, thus helping to set better design targets for navigation lenses.
Freeform technology is one of the most promising solutions to enhance the performance of optical systems. Aiming to demonstrate the freedom and flexibility of optical system design, several challenging design examples using freeform surface are presented, including two wide-angle camera lenses adopting even and odd pedal polynomial formulations for symmetric optical surface description, and an anamorphic cinema lens employing non-rotationally symmetric surfaces for capturing a wider aspect ratio and offering a 1.33X squeeze factor on a 4:3 sensor size. Methods and constraints for designing and optimizing the optical systems are discussed. The optical performances of these design examples are analyzed. The results verify the practicability and effectiveness of the proposed formulations and design strategies in the field of freeform imaging optical design.
Consumers nowadays have a higher expectation on cameras beyond the function of taking photos or videos. In the early years, the applications of cameras with lenses were, however, very restricted. A new generation of super wide-angle lenses capture the user’s surroundings in full 360°, allowing the user’s friends or families to step inside our universe and experience it, live. However, for such wide angle lens coverage, the relative illumination of the lens must take the vector nature of the light into account. Consequently, the polarization of light becomes a critical parameter. Custom coatings are commonly spread on the lens surface to avoid light transmission polarization falloff. But for consumer applications, complex coating is not a practical solution because it raises cost. In this paper, we present how to take care of the polarization during the design and analysis of the lens design. Additionally, design tips are proposed to suppress the polarization impact.
Over the past decades, advances in metamaterial and metasurfaces inspired innovation in many imaging systems. Contrary to conventional optics where ray tracing and a Fourier transform exactly predict light propagation, those approaches are no longer valid in the case of metasurfaces and hybrid glass-metasurface coated lenses. We will discuss in this presentation our latest results and observations to develop new tools that can be incorporated within a lens design software. Analytical solution developed for this implementation will also be discussed. With these new tools, lens design can think of using metasurface within an optimized optical system
We present two miniature all plastic megapixel panomorph lenses for consumer electronics (total track length (TTL) of 6.56 mm) and mobile devices (TTL of 3.80 mm) showing the unique challenges from specification, design, manufacturing and testing phases of these new generation of miniature 180° FoV wide-angle lenses.
In 2014, miniature camera modules are applied to a variety of applications such as webcam, mobile phone, automotive, endoscope, tablets, portable computers and many other products. Mobile phone cameras are probably one of the most challenging parts due to the need for smaller and smaller total track length (TTL) and optimized embedded image processing algorithms. As the technology is developing, higher resolution and higher image quality, new capabilities are required to fulfil the market needs. Consequently, the lens system becomes more complex and requires more optical elements and/or new optical elements. What is the limit? How small an injection molded lens can be? We will discuss those questions by comparing two wide angle lenses for consumer electronic market. The first lens is a 6.56 mm (TTL) panoramic (180° FOV) lens built in 2012. The second is a more recent (2014) panoramic lens (180° FOV) with a TTL of 3.80 mm for mobile phone camera. Both optics are panomorph lenses used with megapixel sensors. Between 2012 and 2014, the development in design and plastic injection molding allowed a reduction of the TTL by more than 40%. This TTL reduction has been achieved by pushing the lens design to the extreme (edge/central air and material thicknesses as well as lens shape). This was also possible due to a better control of the injection molding process and material (low birefringence, haze and thermal stability). These aspects will be presented and discussed. During the next few years, we don’t know if new material will come or new process but we will still need innovative people and industries to push again the limits.
Almost every aspect concerning the design of modern panoramic lenses brings new challenges to optical designers.
Examples of these include ray tracing programs having problems finding the entrance pupil which is moving through the
field of view, production particularities due to the shape of the front lenses, ways of tolerancing these systems having
strong distortion, particular setups required for their characterization and calibration, and algorithms to properly analyze
and make use of the obtained images. To better understand these modern panoramic lenses, the Optical Engineering
Research Laboratory at Laval University has been doing research on them during the past few years. The most
significant results are being presented in this paper.
Controlled distortion, as in commercial panomorph lenses (Immervision), is used to image a specific part of the object
with more pixels than in a normal fisheye lens. This idea is even more useful when a zone of interest vary in time with
dynamically adjustable distortion as in a panoramic locally magnifying imager. Another axis of research is the use of
modern computational techniques such as wavefront coding in wide-angle imaging systems. The particularities of such
techniques when the field of view is large or with anamorphic imagers are considered. Presentation of a novel circular
test bench in our laboratories, required to calibrate and check the image quality of wide-angle imaging system, follows.
Another presented setup uses a laser and diffractive optical elements to compactly calibrate wide-angle lenses. Then, a
discussion of the uniqueness in tolerancing these lenses, especially the front elements due to the large ratio between lens
diameter and entrance pupil diameter, is included. Lastly, particularities with polarization imaging and experiments of
triangle orientation detection tests before and after unwrapping the distorted images are briefly discussed.
We present a lens with a constant total field of view and real-time variable resolution in certain zones
of interest. This smart imaging lens uses an active optical element to modify as desired the local
distortion. This way, while keeping the total field of view constant, the resolution can be increased in a
zone of interest, at the expense of decreasing it somewhere in the remaining part of the field of view.
We first present the concept of this lens, using a deformable mirror as the active surface. Computer
simulations are done with Zemax in which a magnifying power of 2 in a zone of interest representing
10% of the full field of view is achieved, using a f=12.5 mm lens and a F/# of 18. Different
combinations of theses parameters would allow different performances and results. We then present
experimental results of this lens with a prototype built using a ferrofluidic deformable mirror as the
active element. Experimental results of a zone of increased resolution with a magnification of 1.32 and
a zone of decreased resolution with a magnification of 0.80 are obtained.
Panoramic imaging is of growing importance in many applications around the world spurred by the development of
digital imaging. Panoramic lens characteristics are unique and their careful characterization can be a challenge. For
example, the price to pay for a large field of view in this type of lens is high distortion in the image. For vision
applications like security or inspection, a precise knowledge of the distortion introduced by panoramic lenses is essential
to produce natural unwrapped views to the operator. Of special concern is the image quality which must be uniformed
over the entire field of view because all directions are equally important. In addition, two hemispheric images can also
be stitched together to create a complete spherical image. For these reasons, we have developed a dedicated setup to
study the distortion and the image quality produced by panoramic lenses. The test setup is made of a 75-cm radius
cylindrical structure with targets placed on it. Using referenced equally-spaced targets, we obtained the radial image
mapping curves for various azymuthal angles, allowing us to calculate the full-field resolution map. Also, transition
targets were used to find field-dependent spatial frequency where the MTF is 50%. We tested four different panoramic
lenses, two panomorph lenses and two fisheyes. For each lens, we discussed the experimental resolution and MTF
curves and compared some of those results to theoretical design data.
Tolerancing a lens is a basic procedure in lens design. It consists in first defining an appropriate set of tolerances for the
lens, then in adding compensators with their allowable ranges and finally in selecting an appropriate quality criterion
(MTF, RMS spot size, wavefront error, boresight error...) for the given application. The procedure is straightforward
for standard optical systems. However, it becomes more complex when tolerancing very wide angle lenses (larger than
150 degrees). With a large field of view, issues such as severe off-axis pupil shift, considerable distortion and low
relative illumination must be addressed. The pupil shift affects the raytrace as some rays can no longer be traced
properly. For high resolution imagers, particularly for robotic and security applications, the image footprint is most
critical in order to limit or avoid complex calibration procedures. We studied various wide angle lenses and concluded
that most of the distortion comes from the front surface of the lens. Consequently, any variation of the front surface will
greatly affect the image footprint. In this paper, we study the effects on the image footprint of slightly modifying the
front surface of four different lenses: a simple double-gauss for comparison, a fisheye lens, a catadioptric system
(omnidirectional lens) and a Panomorph lens. We also present a method to analyze variations of the image footprint. Our
analysis shows that for wide angle lenses, on which the entrance pupil is much smaller than the front surface,
irregularities (amplitude, slope and location) are critical on both aspherical and spherical front surfaces to predict the
image footprint variation for high resolution cameras. Finally, we present how the entrance pupil varies (location, size)
with the field of view for these optical systems.
We present the research status of a deformable mirror made of a magnetic liquid whose surface is actuated by a
triangular array of small current carrying coils. We demonstrate that the mirror can correct a 11 μm low order aberrated
wavefront to a residual RMS wavefront error 0.05 μm. Recent developments show that these deformable mirrors can
reach a frequency response of several hundred hertz. A new method for linearizing the response of these mirrors is also
presented.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.