FR2762741A1 - CCD array control method - Google Patents

Abstract

The control method can be used for an array (P1 - P25) of photodiodes (Dp). Before being exposed to a light signal during the image gathering phase, the photodiodes are reverse biased by an initial bias voltage having a first value (VA3). A second bias voltage value (Vrp) is applied only to those photodiodes for which the initial reverse bias has been reduced beneath a limit value (vlin) by a very strong exposure during the image gathering phase. During the reading phase which follows the image gathering phase, this avoids the strongly exposed photosensitive points producing strong parasitic currents which may otherwise be detected with the reading of the normally exposed points.

Description

METHOD FOR CONTROLLING AN IMAGE DETECTOR PROVIDING A
GLARE PROTECTION, AND DETECTOR
OF IMAGE IMPLEMENTING THE METHOD
The present invention relates to image detectors using photosensitive dot arrays. It relates to a method for controlling these image detectors, making it possible to limit the effects of overexposure of certain areas of the image. The invention finds a particularly interesting application in the case of the detection of radiological images. It also relates to an image detector implementing this method.

It is common to produce photosensitive dot arrays using thin film deposition techniques of semiconductor materials, for example hydrogenated amorphous silicon (aSiH), on insulating supports. These photosensitive matrices make it possible to detect images contained in visible or near visible radiation. Such photosensitive matrices are also used for the detection of radiographic images; for this purpose it suffices to interpose a scintillating screen between the X-ray radiation and the photosensitive matrix, to convert the radiation
X in light radiation in the wavelength band to which the photosensitive points are sensitive.

 One of the advantages of image detectors using these photosensitive matrices lies in the fact that the images obtained are of the digital type, offering as an advantage in particular ease of processing and storage of the image. However, taking into account the techniques for reading the different photosensitive points, the effects of dazzling due to overexposure of certain areas of the image can have more serious negative consequences than in the case of images obtained by traditional photographic techniques. Indeed in the latter case, the effects of glare remain circumscribed substantially in the overexposed zone or zones, whereas with the photosensitive matrices these effects are reflected at the level of photosensitive points belonging to non-overexposed zones.

 The photosensitive points which form these matrices are organized in rows and columns, and each most often comprise a photodiode, mounted in series with a switching element.

 The switch element can be a so-called switching diode, or a transistor. The "closed" state of the switch element corresponds in the case of a transistor, to the "on" state of the latter (state in which it is put into conduction by a command applied to its gate), and l the "open" state of the switch corresponds to the opposite state of the transistor, that is to say the "blocked" state.

 When it is a diode which fulfills the function of switch: the "closed" state corresponds to the case where a voltage is applied to the diode with the suitable ones to make it lead in the direction of direct conduction, the diode polarities is said then "live polarized"; The "open" state of the switch corresponds to the case in which a voltage is applied to the diode with polarities opposite to those necessary for its conduction in the forward direction, the diode is "blocked", it is then called "polarized" in reverse ". In each photosensitive point, the switching diode and the photodiode are mounted with opposite directions of conduction, in a so-called "head-to-tail" configuration. Such an arrangement is well known, in particular by French patent applications 86 14058 (publication number 2 605 166) and 88 12126 (publication number 2 636 800), in which a matrix of photosensitive dots of the type are described. two diodes in head-to-tail configuration, a method for reading photosensitive points, and a way of making such a photosensitive device.

 FIG. 1 represents a simplified diagram of an image detector 1, comprising a photosensitive matrix 2 of the above-mentioned type, organized in a conventional manner. The matrix 2 comprises photosensitive points P1 to P25, each formed by a photodiode Dp and a switching diode Dc connected in series according to a configuration called "head to tail". The matrix comprises line conductors Y1 to Y5 crossed with column conductors X1 to X5, with at each crossing, a photosensitive point P1 to P25 connected between a line conductor and a column conductor. The photosensitive points P1 to P25 are arranged along lines L1 to L5 of the columns CL1 to CL5.

 In the example of FIG. 1, only 5 rows and 5 columns are represented which define 25 photosensitive points, but such a matrix can have a much greater capacity, being able to go up to several million photosensitive points or pixels.

It should be noted that to detect radiological images, the detector 1 may comprise a layer 9 (symbolized in FIG. 1 by a square in dotted lines) of a scintillating substance deposited on the matrix 2, in order to constitute a scintillating screen converting an incident X-ray in a radiation to which the photosensitive points are sensitive. The matrix 2 can itself be produced on a transparent insulating support 7 (shown in solid lines) in glass for example, against which a so-called additional light source SL (shown in dotted lines) can be pressed, intended to produce a reset to optical level which is further explained in a continuation of the description. This source
SL can be formed for example with light-emitting diodes, as described in a French patent application published with the number 2 598 250. Assuming that the support 7 is located in the same plane as that of the figure, the source SL is located in a deeper plane, opposite the matrix 2 and the scintillator 9.

In each photosensitive point P1 to P25, the two diodes Dp,
Dc are connected together either by their cathode or by their anode as in the example shown. The cathode of the photodiode Dp is connected to a column conductor X there X5, and the cathode of the switching diode Dc is connected to a line conductor Y1 to Y5.

 In an image acquisition or image acquisition phase, the two diodes Dp, Dc of each photosensitive point P1 to P25 are reverse biased, and in this state they each constitute a capacitor. It should be noted that generally the two diodes Dp, Dc are designed so that the capacitance presented by the photodiode Dp is the highest (of the order for example of 50 times). During the image taking phase, charges are produced by the photodiode of each of the photosensitive points PI to P25, charges the quantity of which is a function of the illumination of the photosensitive point.

These charges accumulate at a point "A", on the node (with floating potential) formed at the junction point of the two diodes Dp, Dc.

A read operation makes it possible to know the quantity of charges accumulated by each of the photosensitive points PI to P25. This reading is carried out line by line, simultaneously for all the photosensitive points connected to the same conductor line VI to Y5. To this end, the photosensitive device comprises a line 3 control circuit, whose outputs SY1 to SV5 are respectively connected to the line conductors VI to
Y5.

 The function of the line 3 control circuit is in particular to apply to the various line conductors Y1 to Y5, simultaneously or by successive addressing of each of these line conductors, pulses intended in particular to achieve a polarization of the photodiodes and the reading of the quantities of accumulated charges. Line conductors VI to VS which are not addressed are maintained at a reference potential Vr or quiescent potential, which can be ground for example, and which can also be the same potential as that which is applied to the conductors in column XI to X5.

 Reading the photosensitive points connected to a given line conductor Y1 to VS, generates in each column conductor XI to X5 the circulation of a current proportional to the quantity of charges stored in the corresponding photosensitive point.

The column conductors X1 to X5 are connected to a reading circuit CL, comprising in the example an integrator circuit 5 and, a multiplexer circuit 6 formed for example by a shift register with parallel inputs and serial output which may be of the CCD type (Charge Coupled
Device).

Each column conductor X1 to X5 is connected to a negative input "-" of an amplifier GI to G5 mounted as an integrator. An integration capacity CI to C5 is mounted between the negative input "-" and an output S1 to S5 of each amplifier. The second input "+" of each amplifier G1 to G5 is connected to a potential which in the example is the reference potential Vr, which potential is consequently imposed on all the column conductors X1 to X5. Each amplifier has a switch element II to 15. Each amplifier has a switch element
It says 15 reset (constituted for example by a transistor type
MOS), mounted in parallel with each integration capacity C1 to C5. The outputs SI to S5 of the amplifiers are connected to inputs El to E5 of the multiplexer 6.

This conventional arrangement makes it possible to deliver at the output SM of the multiplexer 6, signals which correspond to the quantities of charges accumulated at points "A" of all the photosensitive points PI to P25. These signals (not shown) are delivered in "series" and line by line (L1 to
L5), that is to say successively for each photosensitive point of the same line and this line after line.

 It should be noted that when reading a photosensitive point, the current flowing in the column conductor X1 to X5 includes not only the current relating to this photosensitive point, but also leakage currents coming from other photosensitive points connected to this column conductor. Most often, these currents which are added to the useful signal can be neglected, but in certain image configurations, the photodiodes can reach and exceed critical polarization levels, which gives these undesirable currents prohibitive values.

 These effects of the polarization levels of the photodiodes will be better perceived with the explanations which follow given, with reference to FIGS. 2a to 2e which retrace steps of a conventional operating control method, applied to the image detector of FIG. 1.

 FIG. 2a represents signals applied by the line control circuit 3 to the first line conductor VI, signals which are therefore applied to all the photosensitive points P1 to P5 constituting the first line L1. FIG. 2d represents variations of a voltage VA, present at point "A" with floating potential of one of these photosensitive points, the first photosensitive point PI for example.

At an instant to which represents the start of an operating cycle, a so-called polarization pulse IP1 (FIG. 2a) is applied to the first line conductor Y1. The two diodes Dp, Dc of each photosensitive point being joined by their anode, the bias pulse IPI is negative with respect to the quiescent voltage Vr, and its amplitude has a value VP1 of the order of, for example, -5V. The polarization pulse IP1 first puts the switching diode Dc in direct conduction, the voltage VA at point "A" (fig. 2d) changes to a value VA1 which corresponds to the elbow voltage of the switching diode Dc, then increases towards a value VA2 corresponding substantially to the voltage
VP1, minus the bend voltage of the switching diode Dc. The photodiode Dp being reverse biased it constitutes a capacitor which is charged at the value VA2.

From an instant t1 when the bias pulse IPI ceases: the voltage on the line conductor VI recovers the rest value Vr and the switching diode Dc is reverse biased and forms a capacitor; ; as a result, the voltage VA at point "A" decreases to a value VA3 by capacitive division. This value VA3 is that of the reverse polarization known as initial of the photodiode Dp, value which is brought to descend for each photosensitive point PI to P25 according to its illumination during an image taking. In the example described, where the polarization pulse IP1 has the sole purpose of fixing an initial polarization value of the photodiodes
Dp, this IPI bias pulse can be applied simultaneously to all line conductors Y1 to VS.

 From an instant t2 when an image taking phase represented in FIG. 2b begins with a slot Phi, each of the photosensitive points P1 to P25 is exposed to a light signal whose intensity is a function of the content of the picture.

 When a light flash or light signal is applied to a photosensitive point, the photocurrent induces the capacity of the photodiode Dp: the reverse bias voltage is reduced and goes from the value VA3 to a value Vax. The light charge Qx detected by this photosensitive point is, at the first order given by the following relation: Qx = (VA3 - Vax) x Cph; where Cph is the capacitance of the photodiode Dp in reverse polarization.

 The electrical characteristics of each photosensitive point are generally optimized so that the value Vax resulting from the exposure is between two voltage limits which define between them a range or expected operating range. The first limit is constituted by the initial polarization value VA3, for the lowest exposure intensity. The second limit Vlin corresponds to the highest expected exposure intensity; it is given by the rest voltage Vr minus a voltage margin Vm. This voltage margin is provided to guarantee that, in a range of exposure intensities corresponding to the expected operating range, the photosensitive points P1 to P25 remain well reverse biased to a value sufficient to avoid non-linearities between the light charge QX and its translation by a voltage variation.

 So therefore from time t2, for the first point PI for example: if this exposure intensity is zero, the voltage VA at point "A" keeps the value VA3 as represented by a curve sI: or at l On the other hand, if this exposure intensity has the maximum expected value, the voltage at point "A" decreases to the limit of the value Vlin of the voltage margin as illustrated by a second curve s2.

 The instant t3 marks the end of the image taking phase Phi. From this instant t3, the voltage VA at point "A" retains substantially the same value until an instant t4 which is the start of a PHL reading phase.

At time t4, the PHL reading phase begins with the application to the first line conductor VI (Fig. 2a), of a reading pulse IL having the same sign (negative) as the bias pulse IP1 (a such read pulse is applied to only one line conductor at a time). In the classic example which is described, the reading pulse IL has an amplitude
VP2 greater than that (VP1) of the polarization pulse IP1, in order to generate so-called "training" charges which, in each photosensitive point, are added to the quantity of charges produced by the photodiode Dp, as explained in the patent documents cited above. These training charges added to the charges corresponding to the useful light signal, make it possible to minimize the harmful effect (particularly at low values) produced by the mediocre qualities which a switching diode Dc exhibits as a switch.

 The application of the reading pulse IL has the effect of putting the switching diode Dc in direct polarization: the latter charges the capacitance constituted by the photodiode Dp. The charge of the photodiode capacitance determines a current on the column conductor XI, while the voltage VA at point "A" increases to a value VA4 which corresponds to the amplitude VP2, minus the elbow voltage of the diode of Dc switching.

 The difference between the amplitude VP1 of the bias pulse IPI and the amplitude VP2 of the read pulse IL (VP2 - VP1) generates, on the first column conductor X1, a current corresponding to the training load; to this drive current is added a current corresponding to the difference between the value of the voltage at point "A" before the image is taken and after (VA3 - Vax), current which represents the light charge.

 This explanation provided for the first photosensitive point PI is valid for the other photosensitive points of the same line LI, which are read at the same time and which give rise simultaneously to currents on the other column conductors X2 to X5. The reading of the photosensitive points being carried out line by line, the other photosensitive points belonging to the other lines L2 to L5, although also connected to these same column conductors, do not in principle induce current there, as long as the line conductor Y2 to VS to which they are connected does not receive a read pulse similar to the IL pulse.

The instant t5 marks the end of the reading of the first line LI. The PHL reading phase continues with successive readings of the L2 lines,
L3, L4 and L5, for which successively read pulses are applied (not shown) in the same manner as described above. The reading phase ends at an instant t6.

Figure 2e represents the "open" and "closed" states of reset switches II to 15 (shown in fig. 1), switches which only allow the integration operation by amplifiers GI to G5 when they are in the "open" state. FIG. 2e illustrates the fact that these switches 11 to 15 are only in the "open" state during the reading phase.
PHL, and therefore only the currents flowing in the column conductors X1 to X5 during a reading phase are taken into account.

At an instant t7, an upgrading phase shown in FIG. 2c begins with a RAN slot. The purpose of this upgrade is to give the voltages VA at points "A", a value lower than that of the initial value of polarization VA3. Indeed, at the end of the reading phase
PHL, the voltage VA of point "A" of all the photosensitive points PI to P25 displays the voltage VA4 higher than the value of the amplitude VP1 of
The polarization pulse IP1. It is therefore necessary, in anticipation of a following operating cycle, to reduce the voltage VA to a value which allows the conduction in the forward direction of the switching diode Dc, during the application of a following bias pulse. IP2. Such a reduction in the value of the voltage VA is obtained from time t7, by exposing the matrix 2 to additional illumination using the additional light source SL previously mentioned.

 The instant t8 represents the end of the additional illumination; at this instant, the voltage VA has been reduced to a value VA5 lower than the initial value VA3 of polarization, value VA3 which can thus again be given to it by the application of a following pulse of polarization IP2, at an instant t9 the start of a new cycle.

 It should be observed that during the reading of the photosensitive points of a given line L1 to L5, the photosensitive points not addressed but located on the same column as a read photosensitive point, deliver a so-called "leakage" current, which is integrated with the useful signal. For each photosensitive point, this current is due mainly to the leakage current of the photodiode Dp, a current which is low and which in any event generates, during the duration of the frame, that is to say of the reading of all lines L1 to L5, negligible integrated noise.

 However, in certain cases, the image to be detected may include zones of very high luminosity, to the point that in the photosensitive points corresponding to these zones, the amount of charges generated by the overexposure is itself large and can determine the direct conduction of the diodes Dp, Dc of these photosensitive points.

 Such a possibility is illustrated in FIG. 2d, from the instant t2, by a third curve in dotted lines marked s3. The instant t2 marks the start of the image taking phase Phi. Assuming that the first photosensitive point P1 is exposed to a very high light intensity, the charges produced by the photodiode Dp generate a large decrease in the reverse bias voltage at point "A". The voltage VA therefore decreases, until possibly passing through zero, that is to say the resting voltage Vr, then changing sign and increasing in positive to a positive value VA +; the photodiode Dp is then polarized directly. The voltage VA then keeps a positive value with respect to the quiescent voltage Vr until the reading phase PHL where, by the application of a reading pulse IL, it is brought to the negative value VA4 and where the photodiode Dp is again reverse biased.

During the time it is polarized live, a photodiode
Dp provides the column conductor with a much larger parasitic current than the usual leakage current, and this large parasitic current is taken into account at the same time as that of the useful signal delivered by a photosensitive point belonging to a different line, but located on a same column.

 Referring again to FIG. 1, assuming for example that a highly overexposed area of the matrix 2 is constituted by the photosensitive points P13, P14 and P18, P19 belonging to the rows L3, L4 and to the columns CL3, CL4. These photosensitive points or pixels having been strongly overexposed, they are found in direct polarization; consequently they deliver even without being addressed, a strong current on the column conductors X3, X4 to which they are connected. In the case of this example, in the line by line reading phase of the matrix which begins with the first line L1, far from the overexposed area, all the other photosensitive points connected to the column conductors X3, X4, although not overexposed, see themselves "parasitized" by a large current before their own signal: therefore, they appear artificially saturated.

 These configurations can occur frequently, in particular in the case of radiological images, where the doses located in the dynamics of the detector, that is to say which do not saturate the latter, are frequently exceeded by so-called "naked flux" doses; these "naked flux" doses correspond to areas of the detector receiving direct X-radiation from the X-ray generator, having passed through no anatomical area and therefore having undergone no attenuation. It is common for these bare flux doses to be 10 to 100 times greater than the maximum dose for which the detector has been optimized. Under these conditions, the photosensitive points located in the image parts receiving these high doses exhibit the malfunctions explained above, and generate intolerable image artifacts, even in the non-overexposed areas.

 The object of the invention is to provide a solution to this problem of the excessive light intensity of one or more areas of the image. It proposes a method for controlling the operation of an image detector, making it possible to eliminate the excess charges produced by these excess light intensities, before these charges disturb the reading of the photosensitive matrix.

 According to the invention, a method of controlling an image detector comprising a matrix of photosensitive dots each comprising a switch element in series with a photodiode, the method consisting on the one hand, in applying a so-called initial bias voltage having a first value to the photodiodes so as to give them a reverse polarization, before proceeding to an image-taking phase during which the matrix is exposed to a light signal and each photosensitive point produces charges the quantity of which depends on its illumination, the charges accumulated in each photosensitive point generating a variation in the initial bias voltage being exerted in the direction of a reduction in this voltage, the method further consisting in reading the photosensitive points in a reading phase occurring after the image-taking phase, the method being characterized in that it further comprises, between the phase of image taking and the reading phase, to confer a reverse bias having a second value, only to photodiodes whose initial bias voltage has been reduced beyond a given limit value.

 This allows, for example for photosensitive points having been subjected to very strong illuminations having led to a polarization reversal of their photodiode, to repolarize these photodiodes in reverse, without modifying the value of the voltage of the photosensitive points which have been subjected to illumination intensities included in the planned measurement range, and therefore without disturbing their reading.

The invention will be better understood on reading the description which follows, given by way of nonlimiting example, with reference to the appended figures, in which:
- Figure 1 already described shows the diagram of an image detector, which can apply the method of the invention;
- Figures 2a to 2e show the steps of a conventional method of controlling the operation of an image detector;
- Figures 3a to 3e form a timing diagram illustrating the control method of the invention;
- Figure 4 shows schematically an image detector, having a matrix whose photosensitive points use a transistor as a switching element.

 In the nonlimiting example described, the method of the invention applies to the control of the image detector shown in FIG. 1, and it differs from the known method described in FIGS. 2a to 2e in that it comprises a additional step called "repolarization", located between the image taking phase and the reading phase. Consequently, the operations are identical with the two methods until the moment of the start of this repolarization step. It should be noted that compared to a conventional line control circuit, the control device 3 shown in FIG. I also includes a voltage source Il (shown in dotted lines), allowing it to define the amplitude of a signal useful in the additional step according to the invention.

 Figures 3a to 3e describe the different steps of the method of the invention. FIG. 3a represents signals applied by the line control circuit 3 to the first line conductor VI; FIG. 3b represents the image taking phase Phi; Figure 3c shows the RAN upgrade phase; FIG. 3d represents variations of the voltage VA at the point "A" at floating potential of the first photosensitive point PI; FIG. 3e represents the "open" state and the "on" state of the reset switches Il to 15 (shown in fig. i).

 At an instant to which represents the start of an operating cycle, an IPI bias pulse (FIG. 3a) is applied to the first line conductor Y1. This impulse <RTI

At an instant t2 when the image taking phase Phi begins, each of the photosensitive points P1 to P25 is exposed to a light signal whose intensity is a function of the content of the image. As already explained with reference to FIG. 2d, on the one hand, if the illumination of the first photosensitive point PI is included in the expected illumination range, the voltage Va remains between the initial polarization value Va3 (materialistic by the first curve s1 for minimum illumination) and the limit Vlin of the voltage margin (materialized by the second curve s2 for maximum illumination), limit Vlin which represents the value of maximum intensity of expected illumination; on the other hand, if the intensity of illumination of the first photosensitive point PI is greater than the maximum intensity predicted, the charges produced generate at point "A" a variation in the voltage VA such that those decreases beyond the value Vlin, until possibly passing through zero, that is to say the resting voltage Vr, and even possibly changing sign and growing in positive until a positive value VA + (as represented by the third curve s3) , the photodiode
Dp then passing in direct polarization.

 From the instant t3 when the image taking phase ends, the voltage VA retains the value it has acquired during this last phase, until an instant t4 at which begins an operation specific to the process of the 'invention.

The invention indeed proposes to intervene, before the reading phase
PHL, only at the photosensitive points whose illumination was higher than the maximum intensity provided, so as to avoid that the effects of the overexposures of certain photosensitive points are not reflected at the photosensitive points normally lit. To this end, the method of the invention consists in applying to all the photosensitive points P1 to P25, a so-called "repolarization" pulse Irp having, with respect to the rest potential Vr, the same polarity as the polarization pulse IP1, and an amplitude suitable for fixing the potential of point "A" at a value equal to or less than the voltage margin limit Vlin, a limit which is reached for the maximum intensities provided.

 According to the invention therefore, a repolarization pulse Irp (shown in FIG. 3a) is applied at time t4 to all the line conductors VI to Y5. The repolarization pulse Irp acts on the voltage VA of all the points "A" of the photosensitive points P1 to P25, according to the value reached by this voltage at the end of the image taking phase Phi, by a same manner as that described below for the first photosensitive point PI.

In the example described, the Irp repolarization pulse is negative. It has an amplitude VP1 'such that, taking into account in particular the bend voltage of the switching diode Dc, it determines at point "A" a so-called repolarization voltage having a value
Vrp included in the voltage margin, that is to say between the rest voltage Vr and the value Vlin of this margin. In these conditions:
- if at the end of the image taking phase Phi, the voltage VA has a value between VA3 and Vlin, that is to say comprised within the range provided, the application of the repolarization pulse Irp remains without effect, since the switching diode Dc having remained reverse biased at a value greater than Vlin, it does not drive; the paces of the first and second curves s1, s2 are not modified.

- if at the end of the image taking phase PHi, the voltage VA has a value between Vlin and the positive value VA +, that is to say resulting from overexposure: in this case, the repolarization pulse Irp has sufficient amplitude to put the switching diode
Cd in direct conduction, and charge the capacitance formed by the photodiode
Dp; the voltage at point "A" then tends towards the value Vrp of the repolarization voltage, value which corresponds to the amplitude of the pulse Irp minus the elbow voltage of the switching diode Dc. We see by the third curve s3 corresponding to the overexposure values, that the voltage VA returns to a reverse polarization value from the instant t4, to reach the repolarization value Vrp at an instant t5.

 The time elapsed between instants t4 and t5 depends on the value of overexposure reached, and the duration T1 of the repolarization pulse Irp must be sufficient to allow the capacitance constituted by the photodiode Dp to be charged at the repolarization value Vrp; for this purpose, the irp repolarization pulse can have substantially the same duration as the IPI polarization pulse. Preferably, the repolarization value Vrp is chosen close to the limit voltage Vlin, so that for the photosensitive points P1 to P25 which have been overexposed, the reverse polarization which is given to them has a value high enough to minimize the current residual leakage.

 The instant t5 marks the end of the repolarization pulse Irp applied to all the line conductors Y1 to VS. From time t5, the voltage VA at point "A" retains the same reverse polarization value until time t6 which is the start of the PHL reading phase.

 At time t6, the reading phase PHL begins with the application to the first line conductor VI (fig. 2a) of a reading pulse IL, in order therefore to read the photosensitive points P1 to P5 of the first line L1 (and only those in the first line). In this example, the read pulse has a greater amplitude VP2 than that VP1 of the bias pulse IPI, in order to create "training" loads as explained above.

As already described with reference to FIGS. 2a to 2e, the application of the read pulse IL has the effect of putting the switching diode
Dc in direct polarization: whatever at this instant the value of the voltage VA, between Va3 and Vlin or else the repolarization value
Vrp, the switching diode Dc charges the capacitance constituted by the photodiode Dp, and consequently brings the voltage VA to the value VA4 which corresponds to the amplitude VP2 of the reading pulse minus the elbow voltage of the switching diode Dc.

 The charge of the photodiode capacitor Dp determines a current on the column conductor Xl. The difference between the amplitude Vp1 of the bias pulse IP1 and the amplitude Vp2 of the read pulse IL generates, on this first column conductor XI, a current corresponding to the drive load. To this drive current is added a current corresponding to the difference in polarization between the value Va3 of the voltage at point "A" before the image is taken, and the value of the voltage at this point just before reading. This last current constitutes the useful current representing the light charge, if the voltage at point "A" is between VA3 and Vlin; but if this difference in polarization is greater because the voltage at point "A" had the value of repolarization Vrp, the measurement is not significant since it is given by an overexposure of the photosensitive point.

 This explanation provided for the first photosensitive point P1 is valid for the other photosensitive points of the same line L1, which are read at the same time, and which give rise simultaneously to currents on the other column conductors X2 to X5.

 The reading of the photosensitive points being carried out line by line, the other photosensitive points belonging to the other lines L2 to L5, although also connected to these same column conductors, do not induce reading current there until they are not themselves in the reading phase. The repolarization step in accordance with the invention makes it possible to eliminate the large parasitic currents liable to be delivered by overexposed photosensitive dots, while photosensitive dots belonging to another line are read. The only parasitic currents that remain are given by the usual leakage currents, the value of which is negligible, and which do not affect the measurements of a significant error.

The instant t7 marks the end of the reading of the first line L1. The PHL reading phase continues with successive readings of the L2 lines,
L3, L4 and L5 and ends at time t8.

FIG. 3e represents the "open" and "closed" states of the reset switches Il to 15 (shown in FIG. 1), switches which authorize the integration operation by the amplifiers Gl to G5 only when they are in the "open" state. FIG. 3e illustrates the fact that these switches 11 to 15 are in the "open" state only during the reading phase
PHL, and therefore only the currents flowing in the column conductors X1 to X5 during a reading phase are taken into account.

 At an instant t9, an upgrading phase shown in FIG. 3c begins with a RAN slot. In the same way as already explained with reference to FIGS. 2a to 2e, the RAN leveling is carried out using the additional source SL. Its purpose is to bring the value of the voltage VA at point "A" to a value VA5 lower than the value VA3 of initial polarization.

 The instant t10 represents the end of the additional illumination, that is to say of the RAN phase; at this instant, the voltage VA has been brought back to a value VA5 lower than the initial value VA3 of polarization, which will make it possible to replace the switching diode Dc in direct polarization, by the application of a following pulse of polarization IP2 in a following operating cycle.

 This description of the process of the invention has been made as a function of an application to the control of the image detector shown in FIG. 1, the photosensitive points of which use switching diodes as switching elements. The switching diodes constitute "switches" of poor quality, which leads to the use of a drive charge technique to improve the reading efficiency when the quantities of charges to be measured are of low value.

 Taking into account the method used to create these training loads, method consisting in applying for reading a voltage (VP2), higher than that (VPI) which is used for the initial polarization, it is indeed necessary for a next cycle to operate a refresher.

But the invention, which consists in interposing a repolarization phase between the image taking and reading phase, can be applied as well if there is no creation of training loads, nor of phase RAN upgrade.

 The invention can also be applied equally well in the case of a method in which the pulse serving for the initial polarization has the same amplitude as that serving for the reading; it should be noted that in such a case, the same pulse can be used both for reading an operating cycle, and for the initial polarization of the following cycle.

 Finally, the invention can also be applied in the case where, in each photosensitive point, the switch function is fulfilled by a transistor, as shown in FIG. 4.

FIG. 4 represents the diagram of an image detector 1 ′ which differs from that shown in FIG. I mainly, in that it comprises a matrix 20 in which, for each photosensitive point P1 ′ to P25 ′,
The switch element connected in series with the photodiode Dp is a transistor
T. In the nonlimiting example described, in each photosensitive point P1 ′ to
P25 ', the transistor T is connected by its source S to the cathode of the photodiode
Dp, that is to say at point "A", its gate G is connected to a row conductor Y1 to VS to which the photosensitive point belongs, and its drain D is connected to the column conductor X1 to X5 to which the photosensitive point belongs. The anodes of all the photodiodes Dp are combined and connected to a voltage source 10, delivering in the example a bias voltage V1-, negative with respect to a reference potential Vr1 which is the ground for example, and to which the conductors columns X1 to X5 are carried.

The "blocked" and "on" states of a transistor T are obtained by applying to its gate G a voltage having, with respect to the source S, an appropriate polarity and value. In the configuration shown in Figure 4, the "blocked" state is obtained (whatever the voltage between the drain
D and the source S of the transistor) by applying to the gate G a so-called blocking voltage -Vg (of the order for example of -10 volts), negative with respect to the reference potential True, this is ie mass in the example. This blocking voltage -Vg is delivered by the outputs SV1 to SVS of the line 3 control circuit, and can constitute the quiescent voltage applied to the line conductors VI to VS:
The "on" state is obtained by applying to the grid G a positive voltage + Vg called conduction voltage, for example of +10 volts relative to True.

This conduction voltage can be delivered by the outputs SY1 to SVS, in the form of a pulse constituting a "passing" state command.

 The method of the invention provides for carrying out, after taking the image Phi and before reading the photosensitive points (as in the case of the detector of FIG. 1), a repolarization operation making it possible to confer on overexposed photosensitive points , a reverse polarization value such that it prevents them from delivering large parasitic currents during the reading of the other photosensitive points.

 To this end, it is proposed in this version of the invention, to control the transistors T so as to put in the "on" state only those of these transistors which have between their source S and their drain D, that is that is to say between the reference voltage Vrl or ground and the point "A" at floating potential, a voltage difference greater than that corresponding to the range of intensity of expected exposure, greater than 3 volts for example. This can be obtained in a simple manner, by applying to the gate G of the transistors T a control pulse called repolarization command Icr (not shown), having an amplitude Vcr less than the value of the blocking voltage -Vg which achieves a complete blocking of the transistor: under these conditions, in fact, the transistor is set to the "on" state if its drain-source voltage exceeds a given value.

So for example, in the case of transistors T whose drain D is at potential (0 volts) of a column conductor X1 to X5, and whose source
S is connected to point "A" of a photosensitive point P1 'to P25': if the voltage at some of the points "A" (and therefore at the cathode of the photodiode) becomes less than -3 volts, the transistors of these points "A" (and only these) are set to the "on" state, if the amplitude Vcr of the pulse applied to the gate G of these transistors is of the order of -8 volts, ie in the nonlimiting example described, 2 volts less than the blocking voltage -Vg. In addition to voltage sources (not shown) which the line 3 control circuit can conventionally include, with a view to delivering signals for setting the "on" and "off" states of the transistors T , it may include a voltage source 12 (shown in dotted lines) allowing it to give the repolarization control pulse the necessary voltage value Vcr.

 The characteristics presented by the transistor T in its "on" state, make it much better able than a diode to fulfill the switch function. Consequently, the use of a transistor in this function dispenses with adding training loads to improve the efficiency of the reading, which makes the operation much simpler.

 Assuming that all the line conductors Y1 to Y5 receive the blocking voltage -Vg (-10 volts relative to the reference voltage True) which then constitutes the resting voltage, an operating cycle can begin with the simultaneous application to all the conductors lines VI to VS, of a first state control pulse passing IcI (+10 volts relative to the reference voltage True). Putting a transistor T in the "on" state has the effect of applying the reference potential Vrl or ground at point "A", that is to say to the cathode of the photodiode Dp whose anode is at the polarization potential V1- negative; it follows that the photodiode is reverse biased, to a value of -5 volts for example, which constitutes its initial reverse bias; it then constitutes a capacity. At time t1 when the control pulse ceases Here, the voltage on the line conductors Y1 to Y5 returns to the value of the blocking voltage -Vg. In all the photosensitive points, the capacitance constituted by the photodiode is then charged to the potential of the reference voltage True, and the photodiode Dp remains reverse biased.

 The photodiode retains the same polarization value until the image taking phase Phi which constitutes the next step.

During the image taking, for each photosensitive point P1 'to
P25 'depending on the intensity of the illumination to which it is exposed, the voltage at point "A" is likely to vary from the value of the reference potential Vrl (corresponding to 0 volts in the example) up to - 3 volts for example for the highest intensity in the expected operating range; for a variation bringing the tension to the point "A" beyond this value, the photosensitive point enters in excess of the tension margin being used to guarantee the linearity as already explained previously, and it can even then if there has been a very high overexposure, reach a more negative value than that (V1-) present on the anode and therefore switch to direct polarization.

 After taking the image, a repolarization control pulse Icr is applied to all the line conductors V1 to VS.

As explained above, this pulse puts in the "on" state only those of the transistors T whose source S (which is connected to the point "A" with floating potential) presents with respect to the drain D a minimum potential difference, which in the example is obtained for the photosensitive points which have gone beyond said voltage margin. In such a case, setting the "on" state of a transistor T repolarizes, for example -3 volts, the corresponding point "A".

 After the repolarization phase, the reading phase occurs during which each line conductor Y1 to Y5 receives, line by line, a second "on" state control pulse lc2 (not shown).

This setting to the "on" state of a transistor has the effect of restoring the reference potential Vrl at the corresponding point "A", which generates on the column conductor X1 to X5 corresponding, the circulation of a current proportional to the potential difference between the reference voltage Vr1 and the potential reached by the point "A" after taking the image.

 It should be noted that the second "passing" state control pulse lc2 was used not only to read the photosensitive points of a given line L1 to L5, but also to carry out the initial reverse polarization of the photodiodes Dp, of so that for a following operating cycle, it is possible to go directly to the image taking phase.

 Of course, the mounting directions of the photodiodes Dp in the above example, and that of the switching diodes Dc with the photodiodes Dp in the previous example, can be reversed if the polarities of the voltages and pulses described are also reversed.

Claims (25)

 1. Method for controlling an image detector comprising a matrix (2, 20) of photosensitive dots (P1 to P25) each comprising a switch element (Dc, T) in series with a photodiode (Dp), the method consisting on the one hand, applying a so-called initial bias voltage having a first value (VA3) to the photodiodes (Dp) so as to give them reverse bias, before carrying out an image taking phase (Phi) during which the matrix (2) is exposed to a light signal and each photosensitive point produces charges the quantity of which depends on its illumination, the charges accumulated in each photosensitive point (P1 to P25) generating a variation in the initial bias voltage s' exerting in the direction of a reduction of this voltage, the method consisting on the other hand of reading the photosensitive points in a reading phase (PHL) occurring after the image taking phase (Phi), the method being char erized in that it further consists, between the image taking phase and the reading phase, of imparting reverse polarization having a second value (Vrp), only to photodiodes (Dp) whose initial reverse polarization has been reduced beyond a given limit value (Vlin).  2. Control method according to claim 1, characterized in that the voltage difference between the first value (VA3) of the initial polarization and the limit value (Vlin) constitutes a range of operation corresponding to the expected exposure intensities.  3. Control method according to one of the preceding claims, characterized in that the second polarization value (Vrp) is equal to or less than the limit value (Vlin).  4. Control method according to one of the preceding claims, characterized in that the switching element is a so-called switching diode (Dc).  5. Control method according to claim 4, characterized in that in each photosensitive point (P1 to P25), the switching diode (Dc) and the photodiode (Dp) are connected in series with opposite directions of conduction compared to each other.  6. Control method according to one of the preceding claims, the photosensitive dots (P1 to P25) being arranged on the one hand, in lines (L1 to L5) each comprising a line conductor (VI to Y5), and on the other share in columns (CL1 to CL5) each comprising a column conductor (X1 to X5), for each photosensitive point (P1 to P25) the switching diode (Dc) being connected to a line conductor and the photodiode (Dp) being connected to a column conductor, the method is characterized in that it consists on the one hand before the image taking phase (Phi), of polarizing the photodiodes (Dp) in reverse to the initial polarization value (VA3) by applying on the line conductors (VI to Y5) a so-called polarization pulse (IP1) having a first amplitude (VPI), and on the other hand between the image taking phase (Phi) and the reading phase (PHL), to apply to these line conductors a so-called polarization pulse (Irp) having a second amplitude (VPI ') lower than the first amplitude (VP1) of the polarization pulse (IP1).  7. Control method according to claim 6, characterized in that a repolarization pulse (Irp) is applied simultaneously to all the line conductors (VI to Y5).  8. Control method according to one of claims I or 2 or 3, characterized in that the switching element is a transistor (T).  9. Control method according to claim 8, the photosensitive points (P1 'to P25') being arranged on the one hand, in lines (L1 to L5) each comprising a line conductor (Y1 to Y5), and on the other hand in columns (CL1 to CL5) each comprising a column conductor (X1 to X5), the method is characterized in that the same first end (anode or cathode) of all the photodiodes (Dp) is connected to a potential (V1-) of a bias voltage, and in that for each photosensitive point (P1 'to P25'), the transistor (T) is connected by its drain (D) to a column conductor (X1 to X5) and by its gate ( G) to a line conductor (Y1 to Y5) and finally by its source (S) to the second end (cathode or anode) of the photodiode (Dp).  10. Control method according to claim 9, characterized in that it consists, in order to polarize the photodiodes (Dp) in reverse to an initial polarization value (Vr1), to apply a pulse to the line conductors (VI to Y5) so-called "on" state control (Ic) having, with respect to the potential of the source (S) of the transistors (T), a polarity opposite to that of a so-called "blocking" voltage (-Vg) of which l application to the gate (G) of a transistor puts the latter in the "blocked" state.  11. Control method according to claim 10, characterized in that the blocking voltage (-Vg) constitutes a rest potential of the line conductors (VI to Y5).  12. Control method according to one of claims 10 or 11, characterized in that it consists, between the image taking phase (Phi) and the reading phase (PHL), to be applied to the line conductors ( VI to Y5) a so-called repolarization control pulse (Icr), having the same polarity as that of the blocking voltage (-Vg) and a value less than that which is necessary for the latter to put a transistor (T) in a "blocked" state.  13. Control method according to claim 12, characterized in that it consists in applying a repolarization control pulse (Icr) to all the line conductors (Vl to Y5) simultaneously.  14. Control method according to one of the preceding claims, characterized in that the image detector being intended for the detection of radiographic images, it comprises a scintillator screen (9) converting X-radiation into light radiation in the band of wavelengths to which photodiodes (Dp) are sensitive.  15. Image detector implementing the method according to any one of claims 1 to 14, comprising a matrix (2) of photosensitive points (P1 to P25), a line control circuit (3), each photosensitive point ( P1 to P25) comprising a switch element (Dc, T) mounted in series with a photodiode (Dp), the photosensitive points being arranged on the one hand, in lines (L1 to L5) each comprising a line conductor (VI to Y5) connected to the line control circuit (3), and on the other hand in columns (CLI to CL5) each comprising a column conductor (X1 to X5), the line control circuit (3) delivering to the line conductors, in a phase which precedes an image taking phase (Phi), a first pulse (IP1,1c) making it possible to give the photodiodes (Dp) a reverse bias voltage known as initial having a first value (VA3), charges liable to be accumulated by each of the photosensitive points (P1 to P25) during the imaging age causing a reduction in the bias voltage, the quantities of charges being read in a reading phase (PHL) occurring after the image taking phase (Phi), characterized in that it further comprises means (3 , 11, 12) to confer, between the image taking phase (Phi) and the reading phase (PHL), a second value (Vrp) of reverse bias voltage only to photodiodes (Dp) whose voltage initial reverse bias has been reduced beyond a given limit value (Vlin).  16. Image detector according to claim 15, characterized in that the second value (Vrp) of reverse polarization is equal to or less than said limit value (Vlin).  17. Image detector according to one of claims 15 or 16, characterized in that it comprises a voltage source (11, 12) cooperating with the line control circuit (3) and the switching elements (Dc, T ), to apply the second reverse bias voltage (Vrp) value to the photodiodes (Dp).  18. Image detector according to one of claims 15 or 16 or 17, characterized in that the switching element is a so-called switching diode (Dc), and in that in each photosensitive point (P1 to P25), the switching diode (Dc) and the photodiode (Dp) are connected in series between a row conductor (Y1 to Y5) and a column conductor (X1 to X5) with opposite directions of conduction with respect to each other.  19. Image detector according to claim 18, characterized in that before the image taking phase (Phi), the line control circuit (3) delivers on the line conductors (Vi to Y5) a so-called polarization pulse (il1), whose amplitude (VP1) makes it possible to define the value (VA3) of the initial reverse polarization, and in that before the reading phase, the line control circuit delivers a so-called repolarization pulse (Irp) of which the amplitude (VP1 ') is lower than that of the polarization pulse (IP1).  20. Image detector according to one of claims 15 or 16 or 17, characterized in that the switching element is a transistor (T).  21. Image detector according to claim 20, characterized in that the same first end (anode or cathode) of all the photodiodes (Dp) is connected to a potential (V1-) of a bias voltage, and in that that for each photosensitive point (P1 'to P25'), the transistor (T) is connected by its drain (D) to a column conductor (X1 to X5) and by its gate (G) to a line conductor (V1 to Y5 ) and finally by its source (S) at the second end (cathode or anode) of the photodiode (Dp).  22. Image detector according to one of claims 20 or 21, characterized in that the line control circuit (3) delivers to the gates (G) transistors (T), ie a voltage called "blocking" (- Vg), the application of which to the gate (G) of a transistor places the latter in the "blocked" state, ie a so-called conduction voltage (+ Vg) having a polarity opposite to that of the blocking voltage and whose application to the gate puts the transistor in the "on" state, or else a voltage (Icr) called repolarization control voltage having the same polarity as that of the blocking voltage (-Vg) and a value (Vcr ) lower than that of the latter, so as to put in the "on" state only transistors (T) whose voltage between the source (S) and the drain (D) is greater than a given potential difference.  23. Image detector according to claim 22, characterized in that the blocking voltage (-Vg) constitutes a quiescent potential of the line conductors (Y1 to Y5).  24. Image detector according to one of claims 22 or 23, characterized in that the voltage (Icr) called repolarization control is applied in the form of a pulse to all the line conductors (Vl to Y5) simultaneously .  25. Image detector according to one of claims 15 to 24, characterized in that the image detector being intended for the detection of radiographic images, it comprises a scintillator screen (9) converting an X-ray radiation into a radiation light in the wavelength band to which the photodiodes (Dp) are sensitive.