JP4149964B2 - Image processing device - Google Patents

Description

  The present invention relates to an image processing apparatus, and more particularly to an image processing apparatus capable of detecting an edge of a document without being affected by disturbance light.

  In an image processing apparatus such as a scanner, it is necessary to detect an edge (edge) of a document to be read. An ADF scanner integrated with an ADF (Automatic Document Feeder) that automatically conveys a document to be read has a structure in which ambient light does not enter the reading unit (image reading unit) during reading. . Therefore, the color of the document backing is physically switched from “white” to “black”, thereby adding a light / dark difference (contrast) with the document to detect the edge of the document.

On the other hand, in a flat bed (FB) type scanner that puts a document to be read on a document placing table one by one, by setting the color of the back surface (the surface in contact with the document) of the document pressing cover to “black”, An edge detection of a document is performed by adding a difference in contrast (contrast) with the document (see Patent Document 1). Alternatively, the color of the back surface of the document pressing cover is detected as a color different from the background color of the document (see Patent Document 2).
JP-A-6-152875 JP 7-264408 A

  As described above, in the FB type scanner, the document pressing cover must be closed as a premise of edge detection. The reason for this is as shown in FIG. That is, when the document pressing cover 15 'is closed, disturbance light does not enter as shown in FIG. Accordingly, as shown in FIG. 3C, the output level of the reading unit 3 ′ is only the detection output of the light reflected from the original 100 ′ by the light from the light source 1 ′. The document 100 ′ can be reliably detected, which is clearly different from the portion. However, when the document pressing cover 15 'is opened, disturbance light G enters as shown in FIG. Therefore, as shown in FIG. 3E, the output level of the reading unit 3 ′ cannot be distinguished from the original 100 ′ because the level of the disturbance light G is close to the level of the reflected light from the original 100 ′. .

  In Patent Document 1 described above, the binary image signal obtained from the read image signal is continuously “white” in order to enable reading of the document with the document pressing cover opened. Detect points. However, the actual disturbance light is extremely unstable light that completely depends on the surrounding environment, and even if it is read by the CCD, it is practically impossible to detect the document (the edge).

  However, as an actual operation for reading an image, it is easier not to close the document pressing cover. Further, when reading a thick medium such as a book, it may be difficult to close the document pressing cover itself. Further, instead of not having to close the document pressing cover, if a pre-scan of the document is required, the reading time becomes long and the operation becomes troublesome, which is not accepted in reality.

  An object of the present invention is to provide an image processing apparatus capable of detecting an edge of a document without being affected by ambient light without closing a document pressing cover.

The image processing apparatus of the present invention includes a light source for reading an image data from a document, a reading unit reads the actual image data is not a pre-scan of the original using the light source, No reading of the image data by the reading unit during the period, for each respective read line, advance the amount of the whole of the light source for the reading line, a first light quantity or off less light intensity than without the first light quantity to perform normal reading a light source control unit to vary any of the second light amount which is determined, on the basis of the image data in the light control unit is read by the reading section while varying the amount of light of the light source, detecting the edge of the document a detection processing unit for, the image data read by the reading unit, and a correction processing section for obtaining an image data that has been corrected based on the variation of the light quantity of the light source in the light control unit Wherein the detection processing unit generates a second image composed of only the second read line read by the second light quantity, it was constituted of only the first reading line read by the first light quantity Generating a first image having a size equal to that of the second image, generating a third image including a difference between the second and first images, a predetermined threshold value and the third image; , And the correction processing unit reads each of the second reading lines read with the second light quantity with the first light quantity before and after that value. The corrected image data is obtained by substituting with a predicted data string composed of values obtained by multiplying the ratio of the average signal level calculated based on the average value of the first reading line .

According to the image processing apparatus of the present invention, during the read of the image data, for each respective read line, the entire amount of light of the light source, to either the first or the second light quantity (intentionally) variations Let At this time, the second light amount is set to a light amount smaller than the first light amount without being turned off. Therefore, there are a first light amount reading line (hereinafter referred to as a first reading line) and a second light amount reading line (hereinafter referred to as a second reading line), and the respective image data can be obtained. Comparing the image data of the first and second reading lines, there is no difference in the image data in the area where the disturbance light is incident (that is, no document exists), but the document exists (that is, the disturbance light is not present). In the image data of the region where the light is not incident, the signal level is proportional to the first and second light quantity levels (see FIG. 3A). Accordingly, the image data of the first and second reading lines are compared, and an area where the signal level changes in proportion to the first and second light quantity levels is detected as an area where the document exists. As a result, the document (the edge) can be reliably detected regardless of the disturbance light. As a result, it is not necessary to close the document pressing cover as an actual operation when reading an actual image that is not pre- scanned, so that the operation of the operator can be facilitated. Further, when reading a thick medium such as a book, it is possible to read it neatly without closing the document pressing cover. Further, since the data for detecting the original can be obtained during the actual image data reading period other than the pre-scan, other operations such as pre-scan of the original are not required, and the burden on the operator is not increased. The reading time does not become longer. At this time, since many reading lines are read with a normal amount of light, edge detection can be performed in a normal reading process, and pre-scanning can be eliminated. Further, since the second reading line is not turned off, a certain amount of image data can be obtained, and a reduction in resolution can be effectively prevented.

  1 and 2 are block diagrams of the image processing apparatus, and show an example of the configuration of the image processing apparatus of the present invention. In particular, FIG. 1 shows a schematic block configuration of the image processing apparatus, and FIG. 2 shows a schematic configuration of a cross section of the image processing apparatus.

  The image processing apparatus includes, for example, an FB scanner, and as illustrated in FIG. 1, as illustrated in FIG. 1, a light source 1, a light source control unit 2 that controls the light source 1, a reading unit 3, and a signal processing unit 4 that processes signals from the reading unit 3. A control unit 5, a memory 6, a memory control unit 7 for controlling the memory 6, a motor 8, a motor control unit 9 for controlling the motor 8, and an image processing unit 10 for processing an image signal processed by the signal processing unit 4. . The image processing unit 10 includes a detection processing unit 11 and a correction processing unit 12. The image processing apparatus reads an image (image data) drawn on a document (paper or medium) 100 by the reading unit 3 and stores the image data processed by the image processing unit 10 in the memory 6. Transmit (transfer) to a personal computer (not shown). The control unit 5 controls the overall processing in such an image processing apparatus. The control part 5 is implement | achieved as a control part which performs the said control by running the control program resident on a main memory (not shown) in CPU (not shown).

  The light source 1, the reading unit 3, and the motor 8 are integrally formed to form an optical unit 13 as shown in FIG. As is well known, the optical unit 13 moves at a predetermined speed in the Y direction (sub-scanning direction) by the motor 8 driven by the motor control unit 9 when reading the document 100. As a result, the reading unit 3 reads an image drawn on the document 100 for each reading line to be described later while moving in the Y direction. In FIG. 2, the motor 8 is not shown. The direction orthogonal to the Y direction is the X direction (main scanning direction).

  The reading unit 3 uses the light source 1 to read image data from the document 100 placed on the document table (FB) 14. The light source 1 is for reading image data from the document 100. The reading unit 3 includes a well-known line sensor including a CCD. The document placing table 14 is a table on which the document 100 is placed, and is made of platen glass as is well known. As shown in FIG. 2A, the light output from the light source 1 is reflected by the original 100 and enters the line sensor. Thereby, the image on the original 100 is read.

  In this example, the light source 1 includes three LED (Light Emitting Diode) light sources 1 of R (red), G (green), and B (blue). The LED that is the light source 1 is provided over the entire area in the X direction (original 100) as shown in FIG. However, actually, the LED that is the light source 1 is provided at one end in the X direction (not shown). The light output from the LED is evenly applied to the entire area in the X direction of the document 100 using a known light guide plate (not shown) provided over the entire area in the X direction.

  In this example, the line sensor as the reading unit 3 is provided over the entire area in the X direction (original 100) as shown in FIG. However, in practice, the line sensor which is the reading unit 3 is provided in a part in the X direction. The light output from the LED and reflected by the original 100 is collected using a known mirror and lens (not shown) and is incident on the line sensor.

  Original document cover 15 is originally intended to hold document 100 placed on document table 14 from above and prevent ambient light from entering, so as to distinguish document 100 from other areas. It is. However, in the present invention, the document pressing cover 15 may or may not be closed so as to press the document 100 as indicated by a dotted line in FIG. Actually, the document pressing cover 15 does not need to be closed according to the present invention, and is provided so as to be removable. Therefore, in the image processing apparatus of the present invention, the document pressing cover 15 may not exist (as a result of being removed by the operator) when the document 100 is read (or always).

  The light source control unit 2 converts the light amount (output) of the light source 1 into a normal light amount (first light amount) or other light amount for each reading line during the period of image data reading by the reading unit 3. (Second light quantity). The first light amount is a predetermined light amount (light having a predetermined intensity) that is empirically determined in advance, and is a light amount for performing normal reading. The second light amount is a light amount smaller than the first light amount (n% light amount of the first light amount, n is a positive integer smaller than 100, and the same hereinafter), and is used to obtain an image signal for edge detection. The amount of light. The value of n is determined empirically in advance. The second light amount is, for example, a light amount that is half the first light amount (n = 50).

  By such control, the light amount of the light source 1 is (intentionally) different in the sub-scanning direction when reading actual (not pre-scan) image data. Note that the light amount of the light source 1 does not differ in the main scanning direction during actual reading. For example, in a reading line in the sub-scanning direction, if the light amount of the light source 1 is the first light amount at the time of reading, the reading result is as shown on the left side of FIG. On the other hand, if the light quantity of the light source 1 is halved (set to the second light quantity) at the time of reading in another reading line in the sub-scanning direction, the reading result is as shown on the right side of FIG. Become.

  In FIG. 3A, the output of the reading unit 3 in an area of the original 100 (where the disturbance light G is not incident) is halved (although proportional to the light amount of the light source 1, that is, the intensity of the light source 1). The relationship does not hold in the incident region (without the document 100) (the proportional relationship does not hold). As described above, the edge of the document 100 is detected based on the comparison between the image data of the reading line in which the light amount of the light source 1 is intentionally changed and the image data of the normal reading line read with the normal light amount. Since it is possible to know in advance the position of the reading line where the amount of light is intentionally changed and the amount (fluctuation level) of the reading line, the image data of the reading line should be corrected to obtain the original image data (level). To correct. Thereby, the pre-scan for edge detection can be made unnecessary.

  In this example, as shown in FIG. 4, the light source control unit 2 varies the light amount of the light source 1 so that there is one second reading line 102 for a plurality of first reading lines 101. That is, the reading is basically performed as the first reading line 101, and the second reading line 102 is inserted at a rate of one for a plurality (in this example, five). The first reading line 101 is a reading line (reading unit in the sub-scanning direction) that is read with a first (normal) light amount, and the second reading line 102 is a reading line that is read with a second light amount. . Thereby, since many reading lines are read with a normal light amount, edge detection can be performed in a normal reading process, and pre-scanning can be eliminated. In addition, in the second reading line 102, a certain amount of image data can be obtained although it is incomplete, and the resolution can be effectively prevented from being lowered.

Specifically, as shown in FIGS. 4 and 5 (A), the light source control unit 2 turns each of the three LED light sources 1 of R, G, and B when reading the first reading line 101. The first (normal) light amount is used. Further, when reading a second reading line 102 different from the first reading line 101, each of the two LED light sources B and G is set to a first (normal) light amount, and an R LED light source is used. The light amount is smaller than the first light amount (half the first light amount). Such light quantity control is possible by controlling the lighting (ON) time of the LED to be controlled, or by controlling the current value applied thereto. The time from when the control signal for reducing the light amount (output) to the LED is applied to when the LED enters the state is, for example, on the order of μ seconds, and thus is sufficiently faster than the reading speed. As shown in FIG. 5A, in the reading result of the second reading line 102, the output of the image data read by the R (red) LED of the reading unit 3 is halved as the light quantity is reduced by half. In reading the second reading line 102, each of the two LED light sources G and R may be set to the first light amount, and the B LED light source may be set to a light amount half the first light amount.

  FIG. 5A is a diagram showing the relationship between the arrangement of the LED light source 1 and the reading line, and is a diagram schematically showing the output of the reading unit 3 in accordance with this. 5B to 5D correspond to FIG. 5A (these will be described later).

  In this way, the type of LED that reduces the amount of light more than the first amount of light is other than the G LED in the LED light source 1. This is because G (green) has the greatest influence on the visibility characteristics of the human eye, so that the human feels that the image read with the amount of G light reduced is a significantly deteriorated image. Therefore, when obtaining corrected image data from image data including the plurality of second reading lines 102, the result of reducing the light amount of the LEDs other than the G LED is closer to the image data read with only the normal light amount. And more accurate correction is possible.

  In addition, it is only 1 LED in the LED light source 1 which consists of 3 LED that a light quantity is made into the normal half. In this example, only the amount of light output by the R LED is half the normal amount of light as shown in FIG. Accordingly, the total light amount for the second reading line 102 is, for example, 5/6 because the light amount output by the B and G LEDs remains normal. FIG. 3A is a diagram illustrating the principle of the present invention for convenience of explanation.

  The image data read by the reading unit 3 is amplified and waveform-shaped by the signal processing unit 4 and then temporarily written in a predetermined area (buffer memory) of the memory 6 by the memory control unit 7 in units of reading lines. Thereafter, the image processing unit 10 reads out the image data stored in the memory 6 via the memory control unit 7 in units of reading lines, and performs various image processing. For example, the image processing unit 10 performs shading correction processing and gamma (Γ) correction processing on the image data, and thereafter performs edge detection processing by the detection processing unit 11 and image correction processing by the correction processing unit 12 described below. In addition, various known image processing is performed as necessary.

The detection processing unit 11 detects the edge of the document 100 based on the image data read by the reading unit 3 while changing the light amount of the light source 1. This image data is actually image data that has undergone the above-described image processing in the image processing unit 10 as described above. The image data is, for example, as shown in FIG. That is, in the entire image 200, there is an area 201 where the document 100 is read at an arbitrary position, and for example, there is an area 202 where ambient light is read on the outer periphery of the area 201. Then, a region 201 that has read the document 100, seen including a plurality of second read line 102 the read region 203 A, a region 202 that has read the ambient light, read a plurality of second read line 102 region 203B and 203C are included. The regions 203A to 203C are referred to as regions 203 obtained by reading a plurality of second reading lines 102.

If this image (image data) is printed out as it is, it looks as follows. That is, since the amount of light than other portions of region 203 A region 201 in a region 201 that has read the document 100 is reduced slightly darker impression of the image, regions 201, regularly for a plurality of read lines (periodic) a darker (hatched) comprises stripes 203 a, it becomes striped feel light and dark. The area 202 from which ambient light is read is, for example, an image in which the background around the document 100 is read. Since the original 100 does not exist in the area 202, there is no influence of a decrease in the amount of light, and the area 203C obtained by reading the plurality of second reading lines 102 in the area 202 does not form a striped pattern as indicated by only dotted lines. The same applies to the region 203B at both ends of the region 203A because it is not included in the region 201. The region 203C is a region that is not included in the region 201 over the entire width of the region 202.

The detection processing unit 11 generates first and second images 204 and 205 as shown in FIG. The first image 204 includes only the first reading line 101 (normal brightness) read with the first light amount, and is an image having the same size as the second image 205. The second image 205 is composed of only the second reading line 102 that is read with the second light amount (slightly dark). The second image 205 is generated by extracting only image data (regions 203 </ b> A to 203 </ b> C) obtained by reading the second reading line 102 from the entire image 200. In the first image 204, two first reading lines 101 before and after the second reading line 102 (regions 203A to 203C) (just before and after, the same applies hereinafter) are extracted, and an average value thereof is calculated. It is generated by doing. Accordingly, the length in the Y direction (sub-scanning direction) of the first and second images 204 and 205 is 1/5 of the entire image 200 in this example. The length in the X direction (main scanning direction) remains unchanged.

  As illustrated in FIG. 6C, the detection processing unit 11 generates a third image 206 including the difference between the first and second images 204 and 205. Here, as can be seen from FIG. 3A, in the third image 206, the area 207 where the document 100 is read and the area 208 where the ambient light around the periphery is read are as follows. That is, if the second reading line 102 is read by reducing the light quantity of the R LED by half, there is not much difference between the signal levels of B and G in the region 207. Further, in the region 208, it is disturbance light and there is not much difference in signal level. Accordingly, the area 207 is substantially equal to a slightly dark image (shown by hatching) read using only the R LED and only a half amount of light, and the area 208 is substantially “black (the gradation level is almost eliminated)”. (Shown with dense diagonal lines).

The detection processing unit 11 detects the edge of the document 100 based on the third image 206. As described above, the first to third images 204 to 206 are images obtained by extracting 1/5 reading lines from the entire image 200. Accordingly, as shown in FIG. 6D, the restored image 209 is obtained by expanding the third image 206 five times in the Y direction (sub-scanning direction) (resolution conversion). The restored image 209 has the same size as the original whole image 200. Since the restored image 209 is obtained by expanding the third image 206 by 5 times, the area 210 where the original 100 is read is slightly dark in the restored image 209, but the edge of the original 100 becomes a clear image, and the outer periphery is disturbed. The area 211 from which the light is read is almost black. That is, in FIG. 3C, it is equal to the result of halving the output of the document portion. This is almost the same as the image read with the document cover 15 having the surface (backing) in contact with the document 100 “black” closed with respect to the edge of the document 100.

  The level of the image signal (output of the original portion) read using only the half light amount using the R LED can be known in advance. Therefore, the edge of the document 100 can be detected by setting the threshold value to, for example, 1/2 of the level of the image signal. That is, the document 100 exists in a portion (pixel position) having a signal level larger than the threshold, and the document 100 does not exist in a portion other than that. Thereby, for each reading line, the position where the document exists is determined in units of pixels, and the edge of the document 100 is determined.

  Actually, since the normal document 100 is rectangular, as the edges of the document 100, inclinations of four straight lines (or expressions representing the straight lines) defining the edges are obtained as is well known. Four intersections where the straight lines intersect are obtained. These pieces of information are stored in a predetermined area of the memory 6. The first to third images 204 to 206 are discarded.

  Note that the detection processing unit 11 may perform edge detection by other means without forming the first image 204 or the like. For example, since the position of the second reading line 102 read with the second light amount is known, the detection processing unit 11 detects the image data of the second reading line 102 and the first (or after) first (or subsequent) image data. A difference from the image data of the reading line 101 may be obtained. As a result, in substantially the same manner as described above, the position where the document exists is determined in units of pixels for each second reading line 102, and the edge of the document 100 is determined.

Further, if the second reading line 102 is read with the light quantity of the R LED reduced by half, the detection processing unit 11 reads the R image data in the image data for the second reading line 102 and the previous ( Alternatively, a difference from the R image data in the image data for the first reading line 101 may be obtained. In this case, the edge detection process can be performed at high speed, but if the original 100 contains a lot of “blue”, the edge may not be detected.

  By the way, in the present invention, the amount of light from the light source 1 is reduced during normal reading of the document 100 in order to eliminate the need for pre-scanning. Therefore, as described above with reference to FIG. 6A, when the image data obtained by this reading is printed out as it is, stripes that partially darken the image appear. Therefore, it is necessary to correct the image data obtained by this reading.

  The correction processing unit 12 obtains corrected image data by correcting the image data read by the reading unit 3 while changing the light amount of the light source 1 based on the change in the light amount of the light source 1. This image data is actually image data that has undergone the above-described image processing in the image processing unit 10 as described above. In this example, the correction processing unit 12 reads the image data of the first reading line 101 with the first (normal) light amount, so that the image data of the second reading line 102 is not changed. Since the reading is performed with the second light amount, the complementary processing is performed as follows.

That is, since the position of the second reading line 102 is known, the correction processing unit 12 obtains corrected image data by replacing each of the second reading lines 102 with a corresponding prediction data string. The corresponding prediction data string is calculated by multiplying the value of the image data of the second reading line 102 (the gradation value for each pixel) by the ratio m of the average signal level. As the average signal level ratio m, an average value (second average value) of all signal levels (gradation values) of the second reading line 102 is obtained, and the first and the first reading lines before and after the second reading line 102 are obtained. An average value of all signal levels (gradation values) of each of the reading lines 101 is obtained, and an average value (first average value) of the two average values is obtained, and m = (first average value) / It is calculated by calculating (second average value).

In the second reading line 102, the light amount of the R LED is only halved, and the image data read using the R LED is not “0”, but as a value (gradation) read with the light amount. Exists. On the other hand, the image does not change so much between the second reading line 102 and the first reading line 101 before and after that, and as a result, the ratio of the signal level (the average value of the gradation) between them is the LED. It can also be considered that it corresponds to the difference in light quantity of the light source 1. Therefore, in order to correct the difference in the amount of light, an average signal level ratio m between them is used. By such processing, the read image data can be reflected in the corrected image data as much as possible.

  In this way, the correction processing unit 12 obtains corrected image data in which the stripes are erased from the read image including stripes that are slightly dark, and then performs the following image processing on the corrected image data. That is, as is well known, the corrected image data is rotated to form an upright image (parallel to the sub-scanning direction) using the inclinations of the four straight lines and the four intersections as a result of the edge detection described above. Then, the image data in the outer peripheral area of the document 100 is deleted. That is, only the area of the original 100 is cut out to obtain final corrected image data.

  Note that the correction processing unit 12 may replace each of the second reading lines 102 with a data string other than the predicted data string. For example, corrected image data may be obtained by replacing each of the second reading lines 102 with a corresponding average value data string. The average value data string includes average values of the image data of the first reading line 101 before and after the second reading line 102.

  The image processing unit 10 rewrites the image data that has been subjected to the image processing in the correction processing unit 12 or the like into another predetermined area (another buffer memory) of the memory 6 via the memory control unit 7 in units of read lines. . Thereafter, the interface unit (not shown) reads the image data stored in the memory 6 via the memory control unit 7 in units of reading lines, and transfers it to a personal computer (not shown).

  FIG. 7 is an image reading process flow, and shows the reading process of the document 100 in the image processing apparatus of the present invention shown in FIG.

  The light source control unit 2 initializes the number of reading lines (step S11), and checks whether the reading line is a reading line (second reading line 102, the same applies hereinafter) with the light amount n% (step S12). If the reading line is n%, the light quantity of the light source 1 is set to n% (step S13). If the reading line is not the reading quantity n% (the same applies to the case of the first reading line 101). Sets the light amount of the light source 1 to a normal light amount (100%) (step S14). Thereafter, the reading unit 3 reads image data from the original 100 with the set light amount (step S15), and the signal processing unit 4 performs predetermined signal processing, and then the image data is stored in the memory 6 via the memory control unit 7. Write to. Thereafter, the light source control unit 2 counts up the number of read lines by +1 (step S16), and checks whether or not reading of image data for all the read lines is completed (step S17). When reading for all the reading lines is not completed, the light source control unit 2 repeats step S12 and subsequent steps. Prior to this, the motor control unit 9 drives the motor 8 to move the optical unit 13 in the sub-scanning direction. When the reading for all the reading lines is completed, the reading process is ended.

  FIG. 8 is an edge detection process flow, and shows an edge detection process in the image processing apparatus (the detection processing unit 11) of the present invention shown in FIG.

  The detection processing unit 11 initializes the number of read lines (step S21), checks whether or not the read line to be processed read from the memory 6 is a read line with a light amount n% (step S22), and the light amount n%. In order to obtain the image (image data) of the current reading line (the reading line) as the second image 205 described above, an image buffer for the light amount n% (a predetermined value in the memory 6) is used. Copy to the area (the same applies hereinafter) (step S23). Thereafter, the detection processing unit 11 generates an image (image data) of the reading line having a normal light amount by averaging the image data of the reading line immediately before and after the current reading line (Step S1). In step S24, in order to obtain the first image 204 described above, it is copied to the image buffer for normal light intensity (step S25), and the number of read lines is incremented by +1 (step S26).

  If the reading line is not the light amount n% in step S22, the detection processing unit 11 omits steps S23 to S25 and executes step S26. Thereafter, the detection processing unit 11 checks whether or not the processing for all the reading lines is completed (step S27). If the processing for all the reading lines is not completed, the processing from step S22 is repeated, When the process for the reading line is completed, the process is terminated. Thereby, the first and second images 204 and 205 are obtained.

  FIG. 9 is an edge detection processing flow, showing edge detection processing in the image processing apparatus (the detection processing unit 11) of the present invention shown in FIG. 1, and is executed following the processing flow of FIG.

The detection processing unit 11 initializes the number of read lines (step S31), and for the read line to be processed, for the light amount n% from the image data (first image 204) copied to the light amount normal image buffer. The image data (second image 205) copied to the image buffer is subtracted (step S32), and the difference is copied to another image buffer (step S33) and read in order to obtain the third image 206 described above. The number of lines is counted up by +1 (step S34). Thereafter, the detection processing unit 11 checks whether or not the processing for all the reading lines has been completed (step S35). If the processing for all the reading lines is not completed, the processing from step S32 is repeated, When the process for the reading line is completed, the process is terminated. Actually, since the third image 206 is obtained by the above processing, the detection processing unit 11 detects the edge of the document 100 based on the obtained third image 206 as described above .

  FIG. 10 is an image correction processing flow, and shows image data correction processing in the image processing apparatus (correction processing unit 12) of the present invention shown in FIG.

  The correction processing unit 12 initializes the number of read lines (step S41), checks whether or not the read line read from the memory 6 is a read line with a light amount n% (step S42), and reads a light line with n%. In this case, an image buffer for correction output is used to obtain the above-described corrected image data from data obtained by multiplying the image (image data) of the current reading line (the reading line) by m times (average signal level). (Step S43), and the number of read lines is incremented by +1 (step S44). On the other hand, in step S42, if the reading line is not the light amount n%, the correction processing unit 12 uses the current reading line (reading line) image (image data) as it is to obtain corrected image data. Therefore, the image data is copied to the image buffer for correction output (step S45), and step S44 is executed. Thereafter, the correction processing unit 12 checks whether or not the reading of the image data for all the reading lines has been completed (step S46), and if the processing for all the reading lines is not completed, the steps after step S42 are repeated. When the processes for all the reading lines are completed, the reading process is terminated. Thereby, corrected image data is obtained.

  In the above example, at the time of reading the second reading line 102, each of the two LED light sources R and G (or G and B) is set to the first (normal) light amount, and B (or The LED light source R) has a light amount smaller than the first light amount (the second light amount as a whole), but the light amount at the time of reading of the second reading line 102 is not limited to this.

  For example, as shown in FIG. 5B, when reading the second reading line 102, the light source control unit 2 sets each of the two LED light sources 1 of B and G (or G and R) to the first. And any LED light source 1 of R (or B) may be turned off (n = 0). In this case, the configuration of the light source 1, the reading of the first reading line 101, the ratio and the position of the first and second reading lines 101 and 102, and the like are as described above.

  In this case, the control of the LED light source 1 can be simplified. On the other hand, in the second reading line 102, for example, image data to be read using an R LED cannot be obtained. For this reason, there is a case where the corrected image data obtained by the complement processing is slightly deteriorated.

  Further, for example, as shown in FIG. 5C, the LED light source 1 is composed of three first LED light sources R, G, and B and one second LED light source R, G, or B. You may make it do. The second LED light source is a dedicated light source for edge detection. As can be seen from FIG. 5C, one second reading line 102 is provided for each first reading line 101. Actually, the first reading line 101 and the second reading line 102 are the same reading line, and it can be considered that they overlap each other. The second reading line 102 is a dedicated reading line for detecting the edge of the document 100 (hereinafter also referred to as a dedicated line).

  In this case, the reading of the first reading line 101 is as described above. The light source controller 2 sets the light amount of the dedicated line (second LED light source) to a light amount smaller than the first light amount when reading the second reading line 102. Specifically, the light amount of the dedicated line is half of the first light amount (n = 50) as shown in FIG. 5A, or is turned off as shown in FIG. n = 0). Thereby, the edge of the document 100 can be detected in the same manner as described above. In addition, it is preferable that LED which is a 2nd LED light source is the same color as LED adjacent before or after that from the viewpoint of obtaining the above-mentioned 3rd image 206. FIG. Therefore, in the case of the arrangement as shown in FIG. 5C, the LED outputs light of the same color as B before the dedicated line or R after the dedicated line (that is, B or R).

  As described above, the edge of the original 100 can be detected, and complete image data of the original 100 can be obtained by reading the first reading line 101, so that the complementary processing is unnecessary. On the other hand, in addition to the R, G, and B reading processes for one reading line, the dedicated line reading process must be performed, so the reading speed is reduced by 33 (= 4/3)%.

  For example, as shown in FIG. 5D, a lamp light source 1 may be used instead of the LED light source 1. Some lamp light sources 1 have a fast response speed such as a xenon lamp, but generally have a characteristic that the response speed is slower than that of an LED light source. For example, it takes several tens of milliseconds from applying an OFF signal to turning it off. Therefore, the light quantity level is controlled so as to be continuously changed as shown in FIG. 5D without controlling the light quantity for each reading line or turning it off. As described above, as the lamp light source 1, it is desirable to use a xenon lamp or a lamp that outputs blue light having a short afterglow characteristic.

  In this case, the light source control unit 2 sets the lamp light source 1 as the first light amount when reading the first reading line 101. At the time of reading on the second reading line 102 different from the first reading line 101, the lamp light source 1 is set to a light amount smaller than the first light amount. Then, as illustrated in FIG. 5D, the light amount of the lamp light source 1 is changed so that the plurality of second reading lines 102 are provided for the plurality of first reading lines 101. As described above, since the response speed of the lamp light source 1 is slow, as shown in FIG. 5D, the change in the light amount of the lamp light source 1 is delayed from the decrease in the voltage applied to the lamp light source 1 or the OFF timing. The amount of change (n) is also not constant, but these can be known empirically in advance. Therefore, a plurality of second reading lines 102 can be known based on this. For example, the edge detection can be performed based on the data string from the average value obtained from these and the first reading line 101 before and after the data string. Further, the corrected image data can be obtained based on the amount of change.

  As mentioned above, although this invention was demonstrated according to the embodiment, this invention can be variously deformed within the scope of the gist.

  For example, the size of the original 100 and the type such as the A4 plate may be obtained based on the result of the edge detection described above. Various image processing in the image processing unit 10 including the detection processing unit 11 and the correction processing unit 12 may be executed by a personal computer that is a host device. That is, the image processing apparatus of the present invention may be composed of an image reading apparatus such as a scanner and an image processing apparatus such as a personal computer. The present invention is not limited to an image reading device such as an FB type scanner, but various image reading devices in which disturbance light is incident on a reading unit that reads an image with reflected light from the light source 1, or an image processing device including the image reading device. Can be widely applied. Further, the present invention is not limited to an image reading device such as a scanner, and can be widely applied when disturbance light is incident on a copy or FAX machine.

As described above, according to the present invention, in the image processing apparatus, during reading of actual image data that is not pre- scanned, the entire light amount of the light source is changed without being turned off for each reading line. By using the obtained image data, it is possible to reliably detect the document (the edge thereof) regardless of the disturbance light. As a result, it is not necessary to close the document pressing cover as an actual operation when reading an image , so that the operation of the operator can be facilitated. Further, even when reading a thick medium, it is possible to read it neatly without closing the document pressing cover. Further, other operations such as pre-scanning of the document are not required, the burden on the operator is not increased, and the reading time is not prolonged.

It is an image processing apparatus block diagram. It is an image processing apparatus block diagram. It is an image processing apparatus block diagram. It is an image reading process explanatory drawing. It is an image reading process explanatory drawing. It is an image reading process explanatory drawing. It is an image reading processing flow. It is an edge detection process flow. It is an edge detection process flow. It is an image correction process flow.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Light source 2 Light source control part 3 Reading part 4 Signal processing part 5 Control part 6 Memory 8 Motor 10 Image processing part 11 Detection processing part 12 Correction | amendment processing part 14 Document mounting table 15 Document pressing cover

Claims (4)

A light source for reading image data from a document;
A reading unit that reads actual image data that is not pre-scanned from the document using the light source;
During the reading of the image data by the reading unit, for each reading line, the total light amount of the light source for the reading line is the first light amount for normal reading or the first light amount without turning off. A light source controller that changes the light amount to any one of a predetermined second light amount that is less than one light amount;
A detection processing unit for detecting an edge of the document based on the image data read by the reading unit while changing a light amount of the light source by the light control unit;
A correction processing unit that obtains image data obtained by correcting the image data read by the reading unit based on fluctuations in the light amount of the light source by the light control unit;
The detection processing unit generates a second image including only the second reading line read with the second light amount, and the first processing unit includes only the first reading line read with the first light amount. A first image having the same size as the second image is generated, a third image including a difference between the second and first images is generated, and the third image is compared with a predetermined threshold value. To detect the edge of the document,
The correction processing unit sets each of the second reading lines read with the second light amount to the value of the second reading line, and is an average value of all signal levels of the second reading line. And the first average value that is the average value of the average values of all the signal levels of the first reading lines before and after the second reading line 102 (first average value) / An image processing apparatus, wherein the corrected image data is obtained by replacing with a predicted data string composed of values multiplied by (second average value) . A second reading line that is read with the second light amount, which is different from the first reading line, with respect to the plurality of first reading lines that the light source control unit reads with the first light amount. The image processing apparatus according to claim 1, wherein the light amount of the light source is varied so that the number of the light sources becomes one. The light source is composed of three LED light sources R, G and B,
The light source controller is
At the time of reading the first reading line, each of the three LED light sources R, G and B is set as the first light amount,
At the time of reading on a second reading line different from the first reading line, each of the R and G LED light sources is set as the first light amount, and the B LED light source is set as the second light amount. The image processing apparatus according to claim 1, wherein each of the two LED light sources G and B is used as the first light amount, and an R LED light source is used as the second light amount. The light source comprises a lamp light source that outputs blue light with a short afterglow characteristic,
The light source controller is
At the time of reading the first reading line, the lamp light source is the first light amount,
The image processing apparatus according to claim 1, wherein the lamp light source is set to the second light amount when reading a second reading line different from the first reading line.

Images