1. Technical Field
The present disclosure relates to performance tests of a system compensating trembling movements of which the human body and in particular the hands are afflicted, during operations benefiting from an object being held as still as possible or with a precision as high as possible.
2. Description of the Related Art
Certain image capture apparatus, such as photography apparatus and digital cameras, are equipped with such a system to compensate for trembling movements during the holding of the apparatus, and thus improve the sharpness of blurry images during the capture of an image. These trembling movements are that much greater when the apparatus is light and can be held with a single hand.
Human trembling movements consist of involuntary vibratory movements that are cumulated with voluntary movements. These vibratory movements generally present amplitudes of several tens of millimeters to several millimeters, and can comprise several harmonics from 0.1 Hz to 20 Hz, as well as presenting angular speeds of about 2°/s and/or linear speeds of about 2 mm/s.
Therefore, electromechanical image stabilization systems exist that comprise acceleration and rotational speed sensors to measure trembling movements, as well as actuators to compensate these movements by displacing optical elements or the image sensor of the image capture apparatus. Due to their higher prices and their large bulk, these systems are usually used with high-range and voluminous photo apparatus, such as reflex cameras.
For lower-range apparatus such as “compact” photographic apparatus and those installed on mobile telephones, image stabilization algorithms have been devised that aim to correct the blurry effect in the captured images. This effect results from trembling movements while the apparatus was held. Certain image stabilization systems are based on a detection of the trembling movements. In some systems, the detection of trembling movements is done with the aid of sensors, and in other systems, it is done by the comparison of successive images. Based on the detected trembling movements, image stabilization algorithms apply a correction to the image that aims to limit or negate the blurry effect. Other image stabilization algorithms consist of using an image taken at the required speed and an image taken at a higher speed to produce a corrected image.
A need exists for a performance test of such image stabilization or correction systems, with the aim of development, characterization, and comparison of different existing image stabilization systems. To this end, these systems may be tested by several people. Nevertheless, such tests turn out to be insufficient due to the difficulty of selecting an adequately representative sample of people to obtain conclusive results. Another problem is test reproducibility, in particular to perform comparisons. Indeed, it is impossible for a person to reproduce exactly the same trembling movement, which is involuntary by nature.
Automated systems exist that allow the displacements and the spatial position of an object to be controlled. These systems turn out, in general, to be incapable of generating amplitude and frequency controlled vibrations with a sufficient precision to simulate, in a reproducible manner, human trembling movements, in particular due to a too high inertia and/or response time. In particular, these systems cannot generate frequency vibrations greater than 10 Hz, whereas a human tremble can reach a frequency of 20 Hz. These systems are also incapable of reproducing a vibratory movement combining several frequencies and amplitudes. Moreover, certain systems with an automated arm need to be within a secure enclosure, which is incompatible with the holding of an image capture apparatus that needs to be directed towards a scene in order to perform stabilization tests of the imaging system.
It may therefore be desirable to use a performance test device of an image stabilization system, in particular for an image capture apparatus, that is capable of reproducing one or more combined human trembling movements, with a sufficient precision of both frequency and amplitude.
Embodiments of the disclosure relate to . . . .
Embodiment examples of the disclosure will be described in the following, in a non-limiting manner in relation with the appended drawings among which:
The mechanical ensemble TS comprises a fixed support 1, and mobile supports 2, 3 that are moveable with respect to the support 1. The mobile support 2 is movably mounted in rotation around an axis Y with respect to the fixed support 1. In the example of
The angular position of the mobile support 2 around the axis Y is controlled by the actuator 12, and the angular position of the mobile support 3 around the axis X is controlled by the actuator 13. Moreover, the sensors 22 and 23 each supply an instantaneous position signal allowing the angular position of the supports 2, 3 around their respective rotation axes Y, X to be determined. The group of sensors 24 supplies instantaneous angular speed signals of the supports 2, 3 around their respective axes Y, X. The mobile support 3 is configured to support an image capture apparatus 5 and comprises, to this end, a holder 4 configured to grasp and force the image capture apparatus to move in concert with the mobile support 3.
In the example of
The sensors 22, 23 are for example linear potentiometers supplying a linear displacement value, allowing an angular displacement value of the mobile supports 2, 3 to be deduced from the distance between the measurement point supplied by the sensor and the rotation axis X or Y. The group of sensors 24 comprises for example gyrometers supplying an angular speed value for each rotation axis X, Y. The actuators 12, 13 can be of the linear electrodynamic actuator type, for example of the “Voice Coil Actuator” type, with the advantage of a rapid and precise positioning.
The group of modules ACQM, CTLM, DRVM, of actuators 12, 13, and of sensors 22, 23, 24 forms a control loop wherein the movements of mobile supports 2, 3 are controlled to follow respective variation curves of the angular position. The control can thus be done in such a manner so as to obtain a precision error of less than 0.01° in the positioning of the mobile supports 2, 3, around their rotation axes X, Y. The frequency of the closed control loop, that is to say, the frequency at which the measurement signals SV are supplied by the module ACQM and at which the command signals CMD are supplied by the module CTLM, can be on the order of 100 to 200 times the rate at which the setpoint values SP are supplied by the processor 30. For example, the processor 30 supplies a setpoint position value SP approximately every 10 ms to each actuator 12, 13, whereas the module CTLM calculates a command value CMD every 62.5 μs to each actuator 12, 13. If the vibratory movement to reproduce has a frequency less than or equal to 20 Hz, the processor 30 therefore supplies at least 5 setpoint values SP per vibration period to simulate, which allows for the reproduction of a vibratory movement, such as sinusoidal, with a sufficient precision.
The module CTLM may also be configured to transmit all the position and speed measurements received from the module ACQM to the processor 30. In this manner, using the processor 30, the user can control with what precision each of the movements applied to the mobile supports 2, 3 follows a variation curve of setpoint values.
The processor 30 is configured to allow the user to introduce, for each rotation axis X, Y, characteristics of a vibratory movement, such as one or more vibration frequencies to combine and a movement amplitude for each frequency. From these characteristics, the processor 30 determines position values separated by a certain period, for example 10 ms, of the support 2, 3. The processor 30 can also be configured to receive, for example in a file, sets of position values of the support 2, 3. As an example, these values may be from a measurement, performed upon a person, of trembling movements.
The processor 30 may also be configured to launch image capture operations by the image capture apparatus 5 fixed to the test device, at moments chosen with respect to vibratory movements applied to the supports 2, 3. The image capture operations have chosen characteristics, such as the focus speed, opening, and sensitivity (ISO). The images taken by the apparatus 5 may be transmitted to the processor 30, which stores them. The processor 30 can then analyze the transmitted images to evaluate a blur level and to correlate the blur level obtained with the vibratory movement applied to the image capture apparatus by the test device, during the capture of an image.
According to one embodiment shown by
The modules ACQM, CTLM, and DRVM are for example made from product commercialized by the company NATIONAL INSTRUMENTS. Thus, the module ACQM is for example made from a chassis PXI-1033 incorporating a 6-axis card PXI-7356 and a continuous power supply card PXI-4110. The module DRVM is for example a 4-axis power motor drive MID-7654. The module CTLM can be made with the aid of a PXI-PCIe8361 card, which can be connected to an internal bus PCI of a computer of the PC type. The setpoints SP may be designated with the help of a program installed in the processor 30, such as the program LabView, also commercialized by the company NATIONAL INSTRUMENTS. The test device may thus generate vibrations that may reach a frequency of 20 Hz and an amplitude of 2°.
The test device that has just been described allows for the reproduction of complex vibratory movements, such as that shown in
An embodiment also relates to a device that allows for the detection and storing of a human vibratory movement, in rotation around the two distinct axes that may be perpendicular, and possibly translated in a direction perpendicular to one of these two axes. Thus,
According to one embodiment, the device TMD also comprises an acceleration sensor 43 that supplies a signal STX proportional to an acceleration of the device TMD along the direction X. The signal STX is also processed by the module ACQ1 that supplies periodically, for example every 10 ms, an acceleration measurement to the computer. These measurements allow an integral calculation to determine position values that may be saved in the file TDB, in order to be supplied as setpoint values SP to the test device of
It will clearly appear to the skilled person that the present disclosure is susceptible to various other embodiments and applications. In particular, the disclosure does not only apply to image capture apparatus, but may also be applied to any device that may be equipped with a system aiming to compensate for human trembling movements while an object is being held by one or both hands.
The various embodiments described above can be combined to provide further embodiments. These and other changes can be made to the embodiments in light of the above-detailed description. In general, in the following claims, the terms used should not be construed to limit the claims to the specific embodiments disclosed in the specification and the claims, but should be construed to include all possible embodiments along with the full scope of equivalents to which such claims are entitled. Accordingly, the claims are not limited by the disclosure.
Number | Date | Country | Kind |
---|---|---|---|
10 00320 | Jan 2010 | FR | national |
Number | Name | Date | Kind |
---|---|---|---|
5665919 | Woyski et al. | Sep 1997 | A |
5771406 | Sakamoto et al. | Jun 1998 | A |
7013731 | Szeremeta et al. | Mar 2006 | B1 |
8100017 | Blick et al. | Jan 2012 | B1 |
8408066 | Romero et al. | Apr 2013 | B1 |
20030109992 | Horiuchi et al. | Jun 2003 | A1 |
20040039760 | Hess et al. | Feb 2004 | A1 |
20100014846 | Nishi et al. | Jan 2010 | A1 |
20120111120 | Brand et al. | May 2012 | A1 |
Number | Date | Country |
---|---|---|
2 135 063 | Aug 1984 | GB |
Number | Date | Country | |
---|---|---|---|
20110181741 A1 | Jul 2011 | US |