Constant resolution continuous hybrid zoom system

Abstract

The present invention relates to an optical device for capturing images of wide-angle scenes with a single camera having a continuous panoramic zoom distortion profile. When combined with a processing unit, the hybrid zoom system creates output images with constant resolution while allowing continuous adjustment of the magnification and field of view of the image without interpolation as in digital zoom systems, or without any moving parts as in optical zoom systems.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 615,252, filed January 9, 2018, entitled “Constant Resolution Continuous Hybrid Zoom System,” the entire contents of which are incorporated herein by reference. Background Technology

[0003] Embodiments of the present invention relate to an optoelectronic device for capturing images of wide-angle scenes using a single camera with a continuous panoramic zoom distortion distribution. Instead of using pixel interpolation to maintain the number of pixels in the final image, moving optics to change magnification and reduce the lens field of view (“FoV”), or a combination of two or more cameras with different FoVs, embodiments of the present invention use a distortion distribution with a large, constant magnification in the central region and reduced magnification in the remainder of the field of view. The distortion distribution is designed to reduce pixel interpolation and maintain a nearly constant image resolution.

[0004] Some existing continuous zoom systems use multiple cameras with different FoVs to archive the continuous zoom effect. By combining information captured by two cameras, it is possible to create zoomed-in images without pixel interpolation. However, the necessity of multiple cameras implies trade-offs with image artifacts, cost, power consumption, size limitations, and weight limitations created by fusing images from different cameras. Solutions using only a single camera will not have those trade-offs.

[0005] Existing purely optical zoom systems can change the magnification and field of view of a lens by moving some components inside the optical lens. However, having moving parts inside the optics increases size and complexity. For some applications, such as miniature wide-angle lenses for consumer electronics, the size constraints are too strict to allow the movement of some optical components to create optical zoom.

[0006] On the other hand, existing pure digital continuous zoom solutions apply computational operations to the image to modify the output field of view. This, as a side effect of the need to display at the same output size, creates new pixels from the original pixels at some point during the operation. This process is also known as upsampling, oversampling, or expansion. This can be done through extrapolation, interpolation, or other means. These new pixels calculated from digital zoom do not contain more optical information about the scene than the original image. This computational operation cannot create additional information and is very limited in its ability to increase the quality of the output image.

[0007] A camera with a continuous resolution zoom distortion distribution is needed, along with an associated algorithm that reduces interpolation and maintains a high-quality level of information about the original scene on each pixel. Summary of the Invention

[0008] To overcome all the previously mentioned problems, embodiments of the present invention describe a method using an imager comprising a wide-angle optical lens with a strong magnification variation from center to edge and an image sensor having a plurality of image sensor pixels combined with a processing unit. The resulting continuous hybrid zoom system is capable of outputting an image at a constant resolution while allowing continuous adjustment of the image magnification and field of view, while limiting interpolation created by purely digital zoom systems and limiting the movement of components like those in purely optical zoom systems. In a preferred embodiment of the invention, the continuous zoom system does not include any movable optical elements in the imaging system at all. In another embodiment of the invention, the only movement in the imaging system relates to the autofocus function and may include movement of the image sensor, optical elements, or the entire lens relative to the image plane. This autofocus may utilize a fixed setting or intelligent autofocus, which adapts to the visible or invisible scene content in the output image depending on the selected output image field of view.

[0009] To provide continuous magnification with constant resolution, a wide-angle lens must have a specific distortion distribution. In the central region of the field of view, corresponding to maximum magnification hybrid zoom (or minimum design field of view), the distortion distribution must have a nearly constant magnification to create a constant resolution output with a pixel ratio close to 1:1 between the available pixels of the image sensor and the pixels of the output image. Then, for the field of view of a wide-angle lens larger than the minimum design field of view, the magnification (distortion distribution) decreases to maintain similar image resolution even with the increased output image field of view. At the edges of the wide-angle lens's field of view, the magnification is minimal and defines the maximum design field of view of the continuous hybrid zoom system. Thus, for any chosen output image field of view, the resolution in pixels per degree at the edges of the output image is always close to a 1:1 pixel ratio where the image sensor pixels are located.

[0010] In an alternative embodiment of the invention, instead of a lens having a higher magnification at the center and a lower magnification toward the edges, the digital image from the imager has a zone of maximum magnification in the off-center area of ​​the image, thereby allowing the hybrid zoom area to also be off-center.

[0011] In an alternative embodiment of the invention, instead of having a wide-angle lens (which has a specific distortion distribution such as higher magnification in the center and lower magnification towards the edges), the camera can use any wide-angle lens and bin the pixels to create the same type of effect, such as higher magnification in the center and lower magnification towards the edges, through a processing unit, electronics or other suitable hardware and / or software.

[0012] At the minimum field of view setting, the processing unit or sensor unit can simply perform cropping of the input image to create the output image, since a constant magnification already produces an almost 1:1 ratio between the image sensor pixels and the output image pixels. Using a hybrid zoom setting other than the minimum field of view, there is oversampling of the source image compared to the output image and residual positive distortion. The processing unit can then digitally compress the center of the image to reduce oversampling and reduce the image resolution from the input resolution to the desired output resolution. This compression by the image processing unit gradually becomes more blurred (softer) until the edge of the selected FoV, where the ratio becomes 1:1 through the design of the distortion distribution for wide-angle lenses.

[0013] In some embodiments of the invention, the intelligent binning processing unit may be coupled to an image sensor, or the intelligent binning hardware may be used to pre-compress the central portion of an image before sending it to the processing unit. The intelligent binning image sensor may perform 1x1, 2x2, 3x3, 1x2, 1x3, 2x3, or any other combination of pixel binning desired to reduce the image resolution in selected areas of the image while always limiting the interpolation ratio between the source and output image resolutions. This intelligent binning image sensor allows the processing unit to reduce data bandwidth or the required compression, which is particularly useful where the processing unit would otherwise require time and power to perform the same task. This intelligent binning image sensor is also useful by reducing the size of the transmitted image, allowing for the saving of the amount of useless information transmitted. Ultimately, it may allow for an increase in the camera's frame rate or signal-to-noise ratio.

[0014] In some embodiments of the invention, the camera may be combined with another camera and may also include optical zoom based on moving parts or digital zoom based on interpolation or oversampling. Attached Figure Description

[0015] The above-described invention and the following detailed description of preferred embodiments will be better understood when read in conjunction with the accompanying drawings. For illustrative purposes, the currently preferred embodiments are shown in the drawings. However, it should be understood that the invention is not limited to the precise arrangements and means shown.

[0016] In the attached diagram:

[0017] Figure 1This is a flowchart illustrating the continuous hybrid zoom process;

[0018] Figure 2 This is a schematic diagram illustrating an image captured from a wide-angle lens, which has a constant magnification in the center and then decreases towards the edges;

[0019] Figure 3 This is a graph showing an example magnification based on changes in the field of view;

[0020] Figure 4 It is a graph showing the more general magnification curve according to the change of field of view;

[0021] Figure 5 This is a schematic diagram illustrating how to use a smart binning sensor to reduce the resolution in the oversampled portion of a compressed image; and

[0022] Figure 6 This is an example layout of an optical lens with a large magnification ratio from the center to the edge of the field of view. Detailed Implementation

[0023] The words “a” and “an” as used in the claims and in the corresponding parts of the specification mean “at least one”.

[0024] Figure 1 A flowchart illustrating the entire process for a continuous hybrid zoom system according to the present invention is shown. The first step 100 is to use an imager having an imaging system and an image sensor, the imaging system having a distortion distribution (as referenced) Figure 3(As explained). Imaging systems typically include classic imaging lenses with refractive elements made of plastic or glass, but may also include other optical elements, such as, but not limited to, diffractive elements, mirrors, filters, etc. The imager 100 is used to capture a scene by converting an optical image from the imaging system into a digital image file in step 110 using its image sensor. The image sensor includes multiple image sensor pixels and can be of any type, such as, but not limited to, CCD, CMOS, NMOS, etc. The digital image file has digital image distortion, which has a substantially constant magnification from the center of the field of view to the minimum design field of view and a substantially decreasing magnification from the minimum design field of view to the maximum design field of view. Distortion in the digital image results in the preferred embodiment due to the optical distortion of the imaging system, but may also result in other embodiments due to the intelligent binning of pixels performed by the intelligent binning unit or due to a combination of both optical distortion in the imaging system of the imager and the intelligent binning of pixels by the intelligent binning unit. The intelligent binning unit is located inside the image sensor or in a separate unit, wherein software or hardware receives the raw image from the image sensor and processes it to create the digital image file. The captured digital image is then transmitted to image processing unit 150, within which, in step 120, an output field of view selection unit is used to select an output field of view value between the minimum and maximum design field of view. This selection unit for selecting the output field of view can be of any type, including pre-stored values ​​in the processing unit, manual input from the user, automatic input based on scene content from the algorithm unit, or requirements from the display or application, or any other source for selecting the output field of view or equivalent zoom level. Image processing unit 150 then uses knowledge of the exact digital image distortion from the digital image from imager 100 (whether due to the imaging system or the smart bin image sensor) to process the digital image by performing fisheye dewarping on the digital image in step 130. Fisheye dewarping is used to generate an image without selecting the zoom area for distortion. In some embodiments, depending on the application, the processing of the fisheye-corrected digital image creates a processed image after a known projection. The known projection of any shape includes, but is not limited to, linear projection, equidistant projection, stereoscopic projection, conformal projection, orthographic projection, and any projection defined by polynomial coefficients. During fisheye correction, the processing unit maintains a ratio close to 1:1 between the number of image sensor pixels and the number of processed image pixels in a region generally at the edge of the selected output field of view, although this region may be located elsewhere in the field of view for some specific applications. If the selected output field of view is larger than the minimum design field of view, the image processing unit then crops the field of view to the selected value and adjusts the output resolution of the image in step 140 to create the final processed image. In step 160, the resulting processed image may then optionally be output from the processing unit to a display device or to an algorithm unit.

[0025] Figure 2An example of a digital image captured by an image sensor designed for continuous hybrid zoom, according to an embodiment of the invention, is shown. In a preferred embodiment, the imaging system creates an optical image of the scene in an image plane, and the image sensor is located at that image plane. Rectangle 200 represents the fully digital image captured by the sensor, including an image coverage area 205 created by the imaging system. This coverage area 205 can sometimes be a circle with a wide-angle lens when the horizontal and vertical dimensions of the image sensor are larger than the image size of the imaging lens; however, the exact shape of the image coverage area 205 on the fully digital image 200 can be any shape according to the invention, including rectangles, ellipses, vertically or horizontally cropped portions of circles, etc. Therefore, in some embodiments according to the invention, instead of rectangle 200 representing the fully digital image captured by the sensor, rectangle 220 represents the imaging area with the active pixels of the image sensor, and the digital image file does not have black corners. In both cases where the image sensor is represented by rectangle 200 or rectangle 220, the central area 210 preferably has a constant magnification, which is the maximum magnification value across the entire field of view. Outside the central zone 210, the magnification is lower and decreases with increasing field of view. The image region representing the maximum zoom level 215 is located within the central zone 210, which has an almost constant magnification. Ideally, this image region has a near 1:1 source pixel to output image pixel ratio when the selected output field of view is the minimum design field of view. Within the scene, there are three faces 223, 225, and 235, which would appear to be almost the same size on a conventional camera. Here, since face 223 is located within the region of higher magnification (central zone 210), its image on the sensor is larger than that of face 225 or 235. After processing by the image processing unit, the final output depends on the selected zoom level or the selected output field of view. When the zoom level is at its maximum, the output image 250 represents the output. In this case, due to the high magnification in the center of the original image created by a lens with distortion, almost no interpolation is needed to modify the number of pixels to fit the output resolution compared to pure digital zoom, and face 255 is displayed with a near 1:1 pixel ratio. On the other hand, when the zoom level is at its minimum, output image 260 represents the output. Here, the central region has been compressed by the image processing unit, so the size in the pixels of face 270 is almost equal to the size of faces 265 and 275. The exact compression applied by the processing unit depends on the selected fisheye-corrected projection, which is application-dependent. While the central region is compressed, the edges of the field of view are processed with an almost 1:1 pixel ratio using a number of captured pixels for each displayed pixel, without compression.Here, compression refers to the process of compressing a processed image with a lower number of pixels in the compressed region by using a higher number of pixels from the original digital image file in the region. This process is also known as pixel downsampling, pixel subsampling, or pixel decimation. Any successive zoom levels between the maximum and minimum zoom can be achieved in the same way, always having a pixel ratio close to 1:1 at the edge regions of the output image due to the unique shape of the distortion distribution from the lens. At each zoom level, for the full field of view, the number of pixels in the input image is always greater than 1:1 or close to 1:1. In some embodiments of the invention, the processing unit simultaneously processes a single digital image 200 into multiple processed images 250 and 260 with different selected output field of view values, which would be impossible for a typical zoom system with moving optical elements.

[0026] Figure 3An example graph 300 of the magnification (or distortion) of an imager with hybrid zoom distortion according to the invention, based on changes in field of view, is shown. In a preferred embodiment, digital image distortion results in magnification being maximum in the central region of the image and minimum in the region near the maximum designed field of view. Field of view 310 represents the minimum designed field of view corresponding to the maximum zoom level. For all fields from the center to the minimum designed field of view 310, the magnification value is ideally close to constant, as shown by the plateau 330 of the curve. However, according to the invention, the plateau 330 of the curve is not strictly required, and deviations from a constant plateau are permitted within the scope of the invention. Field of view 320 represents the maximum designed field of view corresponding to the minimum zoom level. At this field of view, the magnification 350 is typically the lowest value in the entire image. At any output field of view 315 located in the region between the minimum designed field of view 310 and the maximum designed field of view 320, the magnification 340 is between the maximum magnification 330 and the minimum magnification 350. In some embodiments of the invention, the minimum design field of view 310 is defined as a portion of the maximum design field of view 320, such that the ratio of field of view 310 / 320 is substantially equal to the ratio of minimum magnification to maximum magnification 350 / 330. In some other embodiments, there is a difference of up to ±10% between these two ratios. As an example, for a lens with a maximum design field of view 320 of 75°, a maximum magnification 330 of 50 pixels / degree, and a minimum magnification 350 of 10 pixels / degree, the ratio of minimum magnification to maximum magnification is 10 / 50, or 1 / 5. Since the ratio of minimum design field of view to maximum design field of view must be equal, we find that in this example, the minimum design field of view 310 is 15°. In some other embodiments of the invention, instead of defining the minimum design field of view from the maximum / minimum magnification ratio, the minimum design field of view 310 is instead defined as a field of view in which the magnification, calculated per pixel per degree, ranges within ±10% of the magnification value at the center of the field of view or the center of the region of interest (when the region of interest is off-center). In some embodiments of the invention, the ratio between the maximum and minimum magnification is at least 2x. Ideally, the ratio between the number of image sensor pixels and the number of image pixels processed in the region at the edge of the selected output field of view is as close to 1:1 as possible. However, in some embodiments of the invention, this ratio can be as high as 2:1 or 1:2 in the region at the edge of the selected output field of view. To obtain a ratio close to 1:1 at all consecutive zoom levels, the magnification value 340 at each output field of view angle 315, denoted by the symbol θ, must satisfy the following condition:

[0027] .

[0028] For example, if the minimum design FoV 310 has a value of 15° and the maximum magnification 330 has a value of 5x compared to the minimum magnification 350, then the magnification 340 at the output FoV 315 at 60° must be greater than the value given by the following equation:

[0029]

[0030] This results in a magnification greater than 1.25x at 60° compared to a minimum magnification of 350 at the maximum design field of view of 75°. In some embodiments of the invention, a deviation of ±25% from the above formula is permitted to account for lens-to-lens manufacturing errors or to account for design decisions resulting in a smoother distortion curve and simpler manufacturing. In this case, the magnification at a given output field of view in the region between the minimum and maximum design field of view results in:

[0031]

[0032] In some other embodiments, instead of satisfying the ±25% condition at every field of view between the minimum and maximum design field of view, the condition can be satisfied only at a plurality of discrete output field of view values, at which the system according to the invention is used. Figure 3 Only a magnification curve diagram according to an exemplary embodiment of the invention is shown, wherein the flat segments 330 and curve 340 are ideal for equidistant fisheye correction (f-θ projection), but other magnification curve diagrams are also possible. For example, in other embodiments of the invention, when the magnification is given as surface magnification rather than linear magnification, the required magnification must follow an equation proportional to the square root of 1 divided by θ, rather than an equation proportional to 1 divided by (over) θ. For this reason, Figure 4 A more general graph is shown.

[0033] Figure 4 A more general magnification profile 400 according to some other embodiments of the invention is shown. In this magnification profile, a defined minimum design field of view 410 is present. Instead of... Figure 3 In the example, the constant magnification curve, specifically the smooth segment 330, can be of any shape in the central region between the center FoV and the minimum design FoV 410, including but not limited to linear lenses, also known as f-tan(θ) projection lenses. (As in...) Figure 2In view 250, the magnification 430 in the central region can be designed such that when the selected output field of view is the minimum design field of view, the desired output view can be output directly without any fisheye correction, because there is no unwanted distortion to be removed by the processing unit. The magnification value at the minimum design field of view 410 can be the maximum magnification of the imager, but this is not a strict requirement in this embodiment of the invention. The magnification curve 400 also has a maximum design FoV value 420, where the magnification value 450 is typically the minimum. Figure 2 In view 260, at the maximum design FoV 420, or at any other field of view value 415 between the minimum design FoV 410 and the maximum design FoV 420, the processing unit then performs fisheye correction to create a fisheye-corrected view. The fisheye-corrected view can be any projection required for application or display, but such that the ratio between the number of image sensor pixels and the number of output image pixels is close to 1:1 in the output field of view. At any other location in the output field of view, the processing unit compresses the original digital image to produce the desired projection. Figure 4 In general, the magnification value 440 at any chosen FoV value 415 is not constrained by a specific equation, but only by the desired fisheye-corrected output view projection, such that a 1:1 pixel ratio condition is observed at at least one location in the chosen output FoV 415.

[0034] Figure 5An example of using an optional smart binning sensor or processing unit as described in some embodiments of the invention is shown. In this image sensor 500, the number of pixels represented by the smallest square is 18x18. This 18x18 sensor is merely an example illustrating the concept, but the idea is the same as that of multi-megapixel image sensors used in many applications. When this smart binning sensor is used in conjunction with a lens having hybrid zoom distortion, it can use binning or not depending on the selected zoom level. At the maximum zoom level, only the central portion of the image is used, and because the magnification from the lens is almost constant in this portion, almost no binning from the sensor is needed. Therefore, all raw pixels are read in the useful area defined by the selected output field of view. At the other extreme, when the hybrid zoom is at the minimum zoom level (meaning the maximum design field of view), there is an almost 1:1 pixel ratio at the edges of the field of view where the magnification is minimal, but the central region is oversampled. In this case, the smart binning sensor can use, for example, pixels 535 in a 1x1 region 530 facing the edges where no oversampling occurs. In the center where oversampling is greatest, nine individual pixels 515 can be binned together in a 3x3 region 510. This smart binning process is applied to limit the number of pixels read from or transmitted from the image sensor to the processing unit, thereby allowing for an increase in the read frame rate on the sensor or a reduction in the bandwidth required to transmit the image. In the middle region between the center and the edge, four individual pixels 525 can be binned together in a 2x2 region 520. In a real sensor or smart binning unit, smart binning is not limited to a 1x1, 2x2, or 3x3 square region, but can also be a rectangular binning of 1x2, 2x3, 1x3, or any other combination, as long as the final image from the smart binning sensor has sufficient resolution at all points to exceed or approach the output resolution of the output image at a selected zoom level within ±25% of the output resolution of the output image. While smart binning is preferably performed as quickly as possible during the capture process, the location of the smart bins at the sensor level is not limited in this invention. Instead of performing intelligent binning in the sensor, intelligent binning can also be accomplished at any time during image capture via any hardware or software process within the intelligent binning unit, before the digital image is sent to the image processing unit.

[0035] Alternatively, in other embodiments of the invention, the intelligent binning sensor itself can be used to generate a highly distorted image with more pixels in the central portion of the FoV compared to the edges, rather than optically using a wide-angle lens with high distortion. Alternatively, the highly distorted raw image can be created by a combination of distortions in the optical lens and the intelligent binning sensor working together. This intelligent binning has the benefit of changing the magnification ratio and position according to external or internal parameters as needed, and even in real time.

[0036] Figure 6 An example layout of an imaging lens with continuous hybrid zoom distortion according to an embodiment of the invention is shown. In this example, the wide-angle lens 600 includes six optical elements 602, 604, 606, 610, 612, and 614, an aperture stop 608, a sensor housing glass 616 which also potentially acts as a filter, and an image plane 618. However, this exact number of elements is not required according to the invention, and the same inventive method can be achieved using more or fewer optical elements. In this example, the maximum full field of view is 180°, as represented at 638 by a vertical ray entering the lens at a 90° angle to the lens axis, but the method according to the invention is compatible with any field of view from very narrow to extremely wide. In this example schematic, the light rays enter the lens from various equally spaced angles between 0° and 90°, numbered 630, 632, 634, 636, and 638, but in a real lens, the light rays enter the lens at all consecutive angles between 0° and the maximum field of view. The light beam from 630 reaches the image sensor at 650, the light beam from 632 reaches the image sensor at 652, the light beam from 634 reaches the image sensor at 654, the light beam from 636 reaches the image sensor at 656, and the light beam from 638 reaches the image sensor at 658. Even though the incoming light beams 630, 632, 634, 636, and 638 are spaced at equal angles, the positions where each light beam reaches the image sensor are not equally spaced. This is because the orientation... Figure 2 The magnification at the edges of the image is higher than that at the center, and the distance between positions 650 and 652 is greater than that between positions 656 and 658. In this illustrated embodiment, lens elements 602 and 614 include aspherical surfaces to help shape the distortion distribution of the imaging lens. However, this is not a requirement of the invention, and in another embodiment, all surfaces may be spherical. Furthermore, other common types of optical surfaces can be used in lens designs to create a desired magnification profile or to improve other optical properties, including, but not limited to, diffractive surfaces, Fresnel surfaces, conical surfaces, cylindrical surfaces, freeform surfaces, holographic surfaces, surfaces with metamaterials, etc. In this embodiment, all optical elements are refractive and are made of glass, plastic, or crystal. However, in some embodiments of the invention, refractive surfaces may also be used to create a desired continuous zoom magnification profile or to improve other optical properties. Finally, in this example embodiment, lens elements 610 and 612 form a doublet lens to improve the color performance of the imaging system. According to the invention, the use of one or more doublet or triplet lenses is possible but not required.

[0037] In some embodiments of the invention, a hybrid zoom system may use multiple cameras to capture images using a lens with continuous hybrid zoom distortion with at least one camera. In this case, the imager includes multiple imaging systems and multiple image sensors that create multiple digital images. Thus, each imaging system may have different parameters, including orientation and position in the scene, the location where the maximum zoom would be the possible maximum magnification, the intensity of the maximum magnification, or the minimum and maximum FoV for hybrid zoom. An image processing unit then receives multiple images from the multiple cameras, each with potential hybrid zoom in its region of interest. In some embodiments of the invention, if desired, the multiple digital images are stitched together before being processed by the processing unit. The processing unit can then fisheye correct the region of interest and adjust the resolution for the output image to the end user, as in the case of a single camera with a continuous zoom system.

[0038] In some other embodiments of the invention, a continuous zoom optical system is combined with digital zoom to create a hybrid system that utilizes the advantages of both continuous zoom and digital zoom.

[0039] In some embodiments of the invention, the image processing unit may apply optional image enhancements before outputting the image. This may include basic enhancements such as contrast, sharpness, noise reduction, white balance, color correction, etc. It may also include more advanced enhancement techniques, including automated computer imaging techniques such as computational imaging, automated image processing, or automated enhancements based on artificial intelligence algorithms. This can be programmed or learned autonomously via a deep learning neural network. An example embodiment of using "AI" to enhance an image is using deep learning to learn 3D information from the captured image and then applying some image blur to objects far from focus. Because wide-angle lenses have a distortion distribution with large magnification changes across the field of view, any movement of the camera will make objects appear larger or smaller, depending on their position in the field of view and their distance from the lens. These changes in the image can then be used by the AI ​​algorithm to measure distances and calculate 3D information. Ultimately, this 3D information can be used to enhance the output in any way desired by the end user.

[0040] In some other embodiments of the invention, the continuous zoom optical camera is used in conjunction with any one of three automatic common settings (i.e., autofocus (AF), auto exposure (AE), and auto white balance (AWB)), a technique commonly referred to as camera 3A correction. These 3A corrections can be applied at the hardware level within the camera, in hardware improvement units that are not part of the camera, in software algorithms, or in a combination of more than one of the above.

[0041] All of the above are figures and examples of specific image distortion transformation units and methods. In all these examples, the imager is not limited to a wide angle and can have any field of view from very narrow to extremely wide. In all these examples, for simplicity, the methods are presented in picture mode, but the methods can also be applied sequentially multiple times to work in video mode. All these examples are not intended to be an exhaustive enumeration or to limit the scope and spirit of the invention. It will be understood by those skilled in the art that changes can be made to the embodiments described above without departing from the broad inventive concept of the invention. Therefore, it should be understood that the invention is not limited to the specific embodiments disclosed, but is intended to cover modifications within the spirit and scope of the invention as defined by the appended claims.

Claims

1. A method for creating a continuous zoom optical system using an imager that creates a digital image file with a variable magnification across a field of view so as to have a ratio close to 1 : 1 between the number of image sensor pixels and the number of output image pixels in a region of a selected output field of view edge, the method comprising: a. creating an optical image of a scene in an image plane using an imager having at least an imaging system and an image sensor; b. converting the optical image into a digital image by the image sensor of the imager, the image sensor comprising a plurality of image sensor pixels, the digital image having a digital image distortion, wherein there is a constant magnification or magnification increase from a center of a field of view to a minimum design field of view corresponding to a maximum zoom level, and a decreasing magnification from the minimum design field of view to a maximum design field of view corresponding to a minimum zoom level; c. receiving, by a processing unit, a selection of an output field of view value between a minimum design field of view value and a maximum design field of view value; and d. processing, by the processing unit, the digital image so as to create a processed image, the processing unit maintaining a ratio between the number of image sensor pixels and the number of processed image pixels close to 1 : 1 in a region of the selected output field of view edge.

2. The method of claim 1, further comprising: e. outputting the processed image.

3. The method of claim 1, wherein a value corresponding to a minimum design field of view is defined such that a ratio between a maximum magnification and a minimum magnification is equal to a ratio between the minimum design field of view and the maximum design field of view.

4. The method of claim 1, wherein the selection of the output field of view value is received from a user.

5. The method of claim 1, wherein the selection of the output field of view value is automatically received from an algorithm.

6. The method of claim 5, wherein the algorithm uses a spatial position and / or orientation of the imager to calculate the selected output field of view.

7. The method of claim 1, wherein a smart binning process is applied to limit the number of pixels read by the image sensor or transmitted to the processing unit.

8. The method of claim 1, wherein the magnification for each output field of view Q in a region between the minimum design field of view and the maximum design field of view is such that:

9. The method of claim 1, wherein the magnification at a given output field of view Q in a region between the minimum design field of view and the maximum design field of view is such that: 。 10. The method of claim 1, wherein the ratio between the number of image sensor pixels and the number of processed image pixels in a region of the selected output field of view edge is up to 2: 1 or 1 :

2. 。 11. The method of claim 1, wherein the imager comprises a plurality of imaging systems and a plurality of image sensors for creating a plurality of digital images.

12. The method of claim 1, wherein the continuous zoom optical system has no movable optical elements. ​ 13. The method of claim 1, wherein the continuous zoom optical system is combined with digital zoom.

14. The method of claim 1, wherein the processing unit simultaneously processes a single digital image into multiple processed images having different selected output field of view values.

15. The method of claim 1, wherein when the received selection of output field of view value is a value corresponding to the minimum design field of view, the processed image is created by the processing unit without fisheye correction to remove distortion.

16. The method of claim 1, wherein the digital image from the imager has a maximum magnification in off-center regions of the image.

17. A continuous zoom optical system using an imaging system to create a digital image file having a variable magnification across a field of view to bring the resolution ratio between image sensor pixel count and output image pixel count in regions of a selected output field of view edge close to 1 : 1, the optical system comprising: a. an imaging system that creates an optical image of a scene in an image plane; b. an image sensor that converts the optical image into a digital image, the image sensor comprising a plurality of image sensor pixels, and the digital image having a digital image distortion, wherein there is a constant magnification or magnification increase from a field of view center to a minimum design field of view corresponding to a maximum zoom level, and a decreasing magnification from the minimum design field of view to a maximum design field of view corresponding to a minimum zoom level; c. an output field of view value selection unit for receiving a selection of an output field of view value between a minimum design field of view value and a maximum design field of view value; and d. a processing unit that processes the digital image to create a processed image, the processing unit maintaining a ratio between image sensor pixel count and processed image pixel count in regions of a selected output field of view edge close to 1 :

1.

18. The continuous zoom optical system of claim 17, wherein the digital image distortion is caused by: optical distortion produced by the imaging system, and / or smart binning of image sensor pixels by a smart binning unit. ​

Images