Determining the transmission quality of an optical unit in a camera system

Abstract

A method for determining the transmission quality of an optical unit in a camera system to draw conclusions about dirt and/or wear in the optical unit and particularly to determine whether the optical unit requires servicing, includes transforming spatially resolved information relating to at least one image from the camera system sectionally using a frequency transformation so that a sequence of transformation coefficients is determined for each section of the at least one image. Each transformation coefficient is a measure of the energy in a specific frequency range. At least one sequence of transformation coefficients having the highest energy values for the highest frequencies is selected. Using the at least one selected sequence a distribution of the frequencies is determined, the distribution of the frequencies is compared with a reference, and the transmission quality of the optical unit is determined using the comparison.

Claims

1. A method for determining a transmission quality of an optical unit of a camera system, the method comprising: transforming spatially resolved information of at least one image of the camera system section by section by using a frequency transformation, to determine a sequence of transformation coefficients for each section of the at least one image, with each respective transformation coefficient being a measure of an energy value in a specific frequency range; selecting at least one sequence of transformation coefficients having highest energy values for highest frequencies; determining a distribution of the frequencies by using the at least one selected sequence comparing the distribution of the frequencies with a reference; repeating the step of transforming spatially resolved information of at least one image of the camera system and the step of selecting the at least one sequence of transformation coefficients several times, for a plurality of images recorded within a predetermined time interval; determining an averaged frequency distribution by using the selected sequences of the transformation coefficients; and comparing the averaged frequency distribution with the reference, and determining the transmission quality of the optical unit by using the comparison.

2. The method according to claim 1, which further comprises providing the spatially resolved information of the at least one image as spatially resolved brightness information.

3. The method according to claim 1, which further comprises providing the transformation coefficients in a respective sequence each sorted in ascending order according to the frequency range.

4. The method according to claim 1, which further comprises selecting at least one longest sequence of transformation coefficients when selecting the at least one sequence of transformation coefficients.

5. The method according to claim 1, which further comprises determining the transmission quality of the optical unit as being inadequate, when a portion of specific high frequencies lies below a predetermined minimal value in the distribution of the frequencies.

6. The method according to claim 1, which further comprises determining the transmission quality of the optical unit as being inadequate, when the distribution of the frequencies in relation to the reference is moved at least by a specific degree to lower frequencies.

7. The method according to claim 1, which further comprises providing the reference as a reference distribution having been determined by using a reference image of an identical camera system, or the reference having been determined based on the reference distribution.

8. The method according to claim 1, which further comprises using the camera system to record a plurality of images, and keeping a scene imaged by the camera system at least substantially unchanged with the recording of the plurality of images.

9. The method according to claim 1, which further comprises: transmitting the sequences of transformation coefficients for the sections of the at least one image within a signal from the camera system to a receiving unit; selecting and filtering from the signal the at least one sequence of transformation coefficients to be selected; and determining the distribution of the frequencies by using the at least one selected and filtered sequence of transformation coefficients.

10. The method according to claim 9, which further comprises providing the signal as a compressed video stream and selecting and filtering the at least one sequence of transformation coefficients to be selected from the compressed video stream.

11. The method according to claim 1, which further comprises determining a need to service the optical unit for at least one of cleaning, polishing or replacement of at least one optical component of the optical unit, based on the determined transmission quality.

12. The method according to claim 1, which further comprises determining at least one of a degree of contamination or a degree of wear of a transparent protective screen of the optical unit of the camera system.

13. The method according to claim 12, which further comprises attaching the camera system to an outside of a vehicle.

14. An evaluation unit for determining a transmission quality of an optical unit of a camera system, the evaluation unit configured: to efficiently select at least one sequence of transformation coefficients, having highest energy values for highest frequencies, from a plurality of sequences of transformation coefficients, having been produced based on spatially resolved information of at least one image of the camera system by using a frequency transformation to determine a sequence of transformation coefficients for each section of the at least one image, with each respective transformation coefficient being a measure of an energy value in a specific frequency range; to determine a distribution of the frequencies using the at least one selected sequence; to compare the distribution of the frequencies with a reference; to repeat the step of transforming spatially resolved information of at least one image of the camera system and the step of selecting the at least one sequence of transformation coefficients several times, for a plurality of images recorded within a predetermined time interval; to determine an averaged frequency distribution by using the selected sequences of the transformation coefficients; to compare the averaged frequency distribution with the reference; and to determine the transmission quality of the optical unit by using the comparison.

15. The evaluation unit according to claim 14, which further comprises a filter element configured to select and filter the at least one sequence of transformation coefficients to be selected from a compressed video stream being transmitted from the camera system to a receiving unit.

16. The evaluation unit according to claim 14, wherein the evaluation unit is configured to determine at least one of a degree of contamination or a degree of wear of a transparent protective screen of the optical unit of the camera system.

17. The evaluation unit according to claim 16, wherein the camera system is attached to an outside of a vehicle.

18. A system, comprising: a camera system and an evaluation unit according to claim 14; the camera system configured: to record one or more images with spatially resolved information; and to transform the spatially resolved information of at least one image of the camera system section by section using a frequency transformation to determine a sequence of transformation coefficients for each section of the at least one image.

19. The system according to claim 18, wherein the camera system has an optical unit with a transparent protective screen having at least one of a degree of contamination or a degree of wear to be determined.

20. The system according to claim 19, wherein the camera system is attached to an outside of a vehicle.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

(1) FIG. 1 a flow chart for schematic representation of the inventive method,

(2) FIG. 2 a system with a camera system and an evaluation unit for carrying out the inventive method according to FIG. 1,

(3) FIG. 3 a diagram which indicates a distribution of the frequencies which were determined using a selection of sequences of transformation coefficients, wherein the sequences of transformation coefficients were generated on the basis of spatially resolved information of at least one image of the camera system using a frequency transformation, and

(4) FIG. 4 a diagram which shows a reference distribution.

DETAILED DESCRIPTION OF THE INVENTION

(5) FIG. 1 shows a flow chart 2 for schematically representing the inventive method for determining a transmission quality of an optical unit 25 of a camera system 24 (cf. FIG. 2). The optical unit 25 has at least one optical component.

(6) With the method, spatially resolved information 4 of at least one image of a camera system 24 is present. The spatially resolved information 4 relates to spatially resolved brightness information of the at least one image.

(7) In association with FIG. 1, the method is described using spatially resolved information 4 of a single image.

(8) With the method, the spatially resolved information 4 of the image is transformed section by section using a frequency transformation 6, so that a sequence 8 of transformation coefficients is determined for each section of the image.

(9) To determine the sequences 8 of transformation coefficients, the spatially resolved information 4 of any section is transformed into the frequency space by means of a Fourier transformation 6 and can then be quantified using quantification coefficients. The coefficients of the resulting matrix are then sorted according to their frequency. In order to sort the coefficients, the matrix is run through in a zigzag fashion along its counterdiagonal. A sequence of transformation coefficients results for each section. The transformation coefficients are present in the respective sequence 8 according to the frequency f.

(10) A respective transformation coefficient is in each case a measure of an energy value in a specific frequency range.

(11) The greater a respective transformation coefficient in a specific frequency range, the greater therefore the portion of this frequency range in the respective section of the image.

(12) Furthermore, at least one sequence 8 of transformation coefficients, having the highest energy values for the highest frequencies f, is selected (selection 10) in particular from the many sequences 8 for the one image. In this example, a number of such sequences 8 of transformation coefficients is selected. For instance, ten sequences of transformation coefficients can be selected, wherein these ten sequences 8 have the highest energy values for the highest frequencies f.

(13) The transformation coefficients are present in a respective sequence 8 sorted in ascending order according to the frequency range.

(14) To this end, one example of quantified matrices is provided, on the basis of which a sequence 8 of transformation coefficients is determined in each case:

(15) $\begin{array}{cc}\left[\begin{array}{cccccccc}78& 11& 8& 3& 0& 0& 0& 0\\ 2& 0& 0& 0& 0& 0& 0& 0\\ 5& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 1& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\end{array}\right]& \mathrm{Matrix}a\\ \left[\begin{array}{cccccccc}92& 12& 6& 3& 0& 0& 0& 0\\ -3& 0& 0& 0& 0& 0& 0& 0\\ 4& 1& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\end{array}\right]& \mathrm{Matrix}b\\ \left[\begin{array}{cccccccc}80& 10& 5& 3& 0& 0& 0& 0\\ 2& 0& 0& 0& 0& 0& 0& 0\\ 4& -3& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\\ 0& 0& 0& 0& 0& 0& 0& 0\end{array}\right]& \mathrm{Matrix}c\end{array}$

(16) On the basis of these matrices, the subsequent sequences 8 of transformation coefficients (sequences a, b, c) can be determined by way of example:

(17) Sequence a: [78, 11, 2, 5, 0, 8, 3, 0, 0, 0, 1]

(18) Sequence b: [92, 12, −3, 4, 0, 6, 3, 0, 1]

(19) Sequence c: [80, 10, 2, 4, 0, 5, 3, 0, −3]

(20) To select the at least one sequence 8 of transformation coefficients which has the highest energy values for the highest frequencies 7, the following can ensue:

(21) Sequence a is longer than sequence b or sequence c, therefore sequence a has a transformation coefficient (namely the last in the sequence) in a frequency range with higher frequencies than sequence b or c. Consequently, sequence a in this example has the highest energy values for the highest frequencies. A frequency portion with the same frequencies is conversely missing in sequences b and c.

(22) Sequences b and c are the same length. In other words sequence b and sequence c in each case have the last transformation coefficient in a frequency range with equally high frequencies.

(23) The amount of a transformation coefficient is a measure of an energy value in a specific frequency range. The amount of the last transformation coefficient of the sequence c is higher than the amount of the last transformation coefficient in sequence b. Consequently, in this example sequence c has the higher energy value for the same highest frequencies (i.e. for the same highest frequency range) of the sequence 8.

(24) With the selection the longest sequences 8 of transformation coefficients are selected. If sequences 8 of the same length are present, the amounts of the transformation coefficients are therefore compared, starting with the last transformation coefficient.

(25) To select the at least one sequence 8 of transformation coefficients, which has the highest energy values for the highest frequencies f, the following can alternatively ensue:

(26) With the selection of the at least one sequence 8 of transformation coefficients, those transformation coefficients which originate from a shared counterdiagonal can be added to a sum in terms of amount.

(27) For the sequences 8 (a, b, c) specified by way of example above, the following sums result by way of example:

(28) Sums of sequence a: [78; 13; 13; 3; 1]

(29) Sums of sequence b: [92; 15; 10; 4]

(30) Sums of sequence c: [80; 12; 9; 6]

(31) The sequence a of transformation coefficients has the most sums. Consequently, sequence a in this example has the highest energy values for the highest frequencies.

(32) The number of sums is the same for sequences b and c. In other words, sequence b and sequence c have in each case the last transformation coefficients, which originate from a shared counterdiagonal, in a frequency range with equally high frequencies.

(33) The last sum of sequence c is greater than the last sum of sequence b. Consequently, in this example sequence c has the higher energy value for the same highest frequency range.

(34) The at least one sequence 8 of transformation coefficients, which has the most sums, can be selected during the selection of the at least one sequence 8 of transformation coefficients. If the number of sums is the same, then the at least one sequence 8 of transformation coefficients can be selected, the last sum of which is the highest.

(35) Using the selected sequences 8, a distribution 14 of the frequencies f (cf. FIG. 3) is determined (determination 12). In order to determine 12 the distribution 14 of the frequencies f, the transformation coefficients of the selected sequences 8 can be added together in terms of amount for the respective frequency ranges, for instance (cf. FIG. 3).

(36) The distribution 14 of the frequencies f is compared with a reference 16 (cf. FIG. 4) (comparison 18), wherein the transmission quality 20 of the optical unit 25 of the camera system 24 is determined using the comparison 18 (cf. FIG. 4). With the comparison 18, the quantification coefficients which were used when determining the transformation coefficients can be taken into account.

(37) FIG. 2 shows a schematic representation of a system 22 with a camera system 24 and an evaluation unit 26. In this example, the system 22 moreover comprises a receiving unit 28.

(38) The camera system 24 comprises an optical unit 25. The optical unit 25 has at least one optical component, in general a number of optical components. An optical component can be a lens and/or a mirror, but also e.g. a protective screen. A protective screen can be used to protect a lens or a mirror from contamination, damage, etc.

(39) The camera system 24 is preferably the camera system cited in FIG. 1. In particular, the system 22 can be used to carry out the method described in association with FIG. 1.

(40) The camera system 24 records at least one image. In this example the camera system 24 records a number of images, in particular in the form of a video.

(41) A component of a technical unit (not shown) is monitored by means of the camera system 24. The camera system 24 is permanently directed at a specific scene with the component to be monitored. The alignment of the camera system with respect to the component does not change.

(42) This can ensure that the scene imaged by the camera system 24 remains at least essentially unchanged during the recording of a number of images.

(43) The camera system 24 is connected to the receiving unit 28 by way of a data link 30.

(44) The recorded images are compressed by the camera system 24 by means of a video compression. The frequency transformation 6 cited in FIG. 1 takes place within the scope of this video compression. For instance, the frequency transformation 6 can be a cosine transformation.

(45) Since this is a video compression, the step of transforming spatially resolved information 4 of at least one image of the camera system 24 is repeated several times within a predetermined time interval.

(46) With the video compression, a compressed video stream is produced, which has the sequences 8 of transformation coefficients cited in FIG. 1.

(47) The compressed video stream is transmitted by the camera system 24 via the data link 30 to the receiving unit 28.

(48) The evaluation unit 26 comprises a filter element 32, which is designed to select and filter the sequences 8 of transformation coefficients to be selected from a compressed video stream, which is transmitted by the camera system 24 to a receiving unit 28.

(49) The step of selecting the sequences of transformation coefficients is repeated several times within the predetermined time interval. For instance, sequences 8 of transformation coefficients can be selected and filtered within the predetermined time interval at different points in time in each instance.

(50) In this example, the filter element 32 is embodied as a proxy and interposed between the camera system 24 and the receiving unit 28.

(51) Furthermore, the evaluation unit 26 comprises a computer unit 34. The computer unit 34 is connected to the filter element 32 by way of a further data link 30.

(52) The selected (and filtered) sequences 8 of transformation coefficients are transmitted by the filter element 32 via the further data link 30 to the computer unit 34 of the evaluation unit 26.

(53) The computer unit 34 of the evaluation unit 26 is designed to determine the distribution 14 of the frequencies f (cf. FIG. 3) using the selected sequences 9, to compare the distribution 14 of the frequencies f with a reference 16 (cf. FIG. 4) and to determine the transmission quality 20 of the optical unit 25 of the camera system 24 using the comparison 18.

(54) For instance, an averaged frequency distribution can be determined for the predetermined time interval by using the sequences 8 selected in the time interval as a distribution 14 of the frequencies f.

(55) An interval of a number of hours or days can be predetermined as a time interval so that an influence of weather conditions and brightness fluctuations due to the time of day is reduced.

(56) FIG. 3 shows a diagram 36, which shows a distribution 14 of the frequencies f by way of example. This can be an averaged frequency distribution.

(57) The distribution 14 of the frequencies f was, as described under FIG. 1 and FIG. 2, determined using a selection of sequences 8 of transformation coefficients, wherein the sequences 8 of transformation coefficients were determined on the basis of spatially resolved information 4 of at least one image of a camera system 24 using a frequency transformation 6.

(58) In order to determine the distribution 14 of the frequencies f, those transformation coefficients of the selected sequences 8, which transformation coefficients apply in each case to the same frequencies, i.e. the same frequency range, were added together in terms of amount.

(59) Since an amount of a respective transformation coefficient is a measure of an energy value E in a specific frequency range, the sum of a number of transformation coefficients in terms of amount is in turn a measure of an energy value E.

(60) In the diagram 36 the frequency f is plotted on the x-axis. The energy value E is plotted on the y axis.

(61) The distribution 14 of the frequencies f is shown as a bar chart.

(62) The highest energy value E is to be registered in the region 42 of the lowest frequencies f (shown to the far left in the drawing).

(63) The distribution 14 of the frequencies f shown by way of example in the diagram 36 is compared with a reference 16 (cf. FIG. 4). In this example the reference 16 is a reference distribution 40.

(64) FIG. 4 shows a diagram 38, which shows a reference distribution 40 by way of example.

(65) The reference distribution 40 can be determined similarly to the distribution 14 of the frequencies f in FIG. 3, but using the spatially resolved information 4 of at least one reference image. The at least one reference image can be for instance the first image(s) recorded by means of the camera system 24 when monitoring the component. In particular, the scene imaged by the camera system 24 in the reference image is the same as in the determination of the at least one image cited in FIG. 1 and FIG. 2.

(66) In the diagram 38, similarly to the diagram 36 in FIG. 3, the frequency f is plotted on the x-axis and the energy value E is plotted on the y axis. The reference distribution 40 is shown as a bar chart.

(67) Here, too, the highest energy value E is to be registered in the region 42 of the lowest frequencies f (shown to the far left in the drawing). It is however to be ascertained that energy values E were determined (shown to the right in the drawing) in a region 44 with higher frequencies f, i.e. in higher frequency ranges, than in the diagram 36 in FIG. 3.

(68) With the comparison of the distribution 14 of the frequencies f from FIG. 3 with the reference 16 embodied as the reference distribution 40 in FIG. 4, it is consequently ascertained that in the distribution 14 of the frequencies f, the portion of high frequencies f, i.e. of frequencies in the region 44, has dropped to zero.

(69) With the comparison, it can be further ascertained that the distribution 14 of the frequencies f in relation to the reference distribution 40 has clearly moved to lower frequencies.

(70) The transmission quality 20 of the optical unit 25 (cf. FIG. 2) is determined as inadequate on the basis of the comparison.

(71) A need to service the optical unit 25, in which for instance at least one optical component of the optical unit 25 is cleaned, polished and/or replaced, is determined on the basis of the determined transmission quality 20. Since the determined transmission quality 20 is “inadequate”, it is ascertained that the transmitting optical components of the camera system 24 require cleaning.

(72) Although the invention has been illustrated and described in detail by the preferred exemplary embodiments, the invention is therefore not restricted by the disclosed examples and other variations can be derived by the person skilled in the art without departing from the scope of the invention.