JP2900495B2 - Image reading method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading method for inputting an image such as a character or a figure by a manual feed operation.

[Prior Art] In a hand-operated image reading apparatus, a conventional reading method using two one-dimensional image sensors includes:
JP-A-60-5383 is known. The method includes arranging a first sensor and a second sensor at regular intervals in the hand feeding direction, and calculating an exclusive OR of an image read at a position where the first sensor is located and an image read by the second sensor. When the degree of coincidence becomes maximum, the second sensor is positioned at the same position as the reading position of the first sensor.
The speed of the first and second sensors is detected from the number of reading scans during that period, and an image without thinning distortion is obtained from the read image based on the speed.

[Problems to be Solved by the Invention] The method detects an identical image of images obtained by two sensors, determines the amount of image thinning based on the number of scans until the same image is read, and uniformly thins the image. To obtain a distortion-corrected image. Such a method of correcting distortion is based on the assumption that the manual feed speed of the image reading device is constant, and when the image input device is actually moved by the manual feed operation to input an image, the speed is constant. It hardly moves. Therefore, distortion remains in the image after the thinning is performed, and it is impossible to obtain a clear image. There is also a method of connecting the points of the detected speed with a continuous curve in order to respond to the speed change. This method is very effective when the actual speed change has a rule because the estimated speed changes continuously. However, when the speed change has no rule and the detection speed is sparse as shown in FIGS. 201 to 209, the estimated speed changes very sharply as shown in FIG. Therefore, when the distortion correction of the input image is performed at such an estimated speed, a new distortion may be generated in the corrected image.

SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide an image reading method for accurately removing distortion from an input image while the hand feed speed changes irregularly.

[Means for Solving the Problems] An image reading method according to the present invention includes at least two of a first sensor and a second sensor, each of which is a one-dimensional image sensor for reading a character or a figure, and has a first sensor and a second sensor. An image read at a position where the first sensor proceeds to
In an image reading method of an image input device that detects a time until the second sensor reads afterward, estimates a scanning speed from the time, corrects a distortion of an input image, and reads an image, the horizontal axis represents time, and the vertical axis represents time. Assuming a speed graph, a function approximating a curve passing through the estimated speed at the N-th time estimated last by the image input device and the detected speed at the time T (N + 1) after the N-th is estimated. An estimated time of the image input device is defined as the function of the estimated speed of the image input device from the 1Nth time to the N + 1th time, where an intermediate time between the Nth time and the T (N + 1) time is the N + 1th time. And the speed of the curve represented by
(Integral numbers of 2, 3,...) Are sequentially increased so that the scanning speed of the image input device is approximated by a curve of the function.

EXAMPLES Hereinafter, the present invention will be described in detail based on examples.

FIG. 1 is a diagram showing a speed change estimated by the method of the present invention. FIG. 2 is a diagram showing a change in speed when the detection speed is connected by a continuous curve by a conventional method. FIG. 3 is a diagram showing a process of estimating a speed change according to the present invention. FIG. 4 shows an image reading apparatus having two sensors arranged in parallel to the manual feeding direction. FIG. 5 is an image obtained from the two sensors input by the image reading device.

Hereinafter, the image reading method of the present invention will be described. The image reading apparatus used in the present invention includes two one-dimensional image sensors (402 and 40) arranged in parallel to the hand feeding direction 408.
3), wherein the two one-dimensional image sensors include:
The reflected light of the image 407 is formed by the range 406. The image 407 is constantly illuminated by a light source (not shown) such as an LED. When the image input device is moved by the manual feeding operation in the manual feeding direction 408 by the image input device and the image is input, the two one-dimensional image sensors (402 and 403) arranged in parallel are connected to 404 and 405. An image of the indicated area is input. The one-dimensional image sensor (40
2 and 403) are arranged in parallel at an interval d401, and the lens 406 is set to project an image one-to-one, so that the interval 409 between the reading positions 404 and 405 is the same as the interval d401. . The hand feed direction of the image input device is 40
8, the image information at the position 405 is input to the sensor 403, and the image information at the position 404 is input to the sensor 402.
03 (hereinafter referred to as the first sensor)
An image of the same point is input to the sensor 402 (hereinafter, referred to as a second sensor) with a delay of the movement time of d401.

Next, a description will be given of a method of detecting characteristic points of an image input by a manual feeding operation by the image input device and correcting distortion. The image written on the manuscript FIG.
When input is performed by a manual feed operation using the image input device shown in the figure, the manual feed speed is not constant, so the first sensor 403 changes to the image shown in FIG. 5 (b), and the second sensor 402 changes to the image shown in FIG. 5 (c). The image shown is obtained. Note that the two images 5
In FIG. 5B and FIG. 5C, the horizontal direction is indicated by a time axis 501, and images at the same time on the time axis 501 are input simultaneously. Therefore, by detecting the time lag of the image at the same point of the two images, the first
The moving speed of the image input device during this interval is determined from the distance d between the sensor and the second sensor, and an image without distortion can be obtained.

Next, a method of estimating the moving speed of the image input device based on the speed will be described. Now, it is assumed that the same image is detected from the two input images and the moving speed of the image input device is obtained, and the results of FIG. 1 (FIG. 2) are 101 to 109 (201 to 209). It should be noted here that detecting the same image and detecting the moving speed of the image input device from the delay of the capture time is the average moving speed of the image input device during the capture delay time. For example, the speed shown in FIG. 101 is the average value of the moving speed of the image input device at a time around the time at which 101 exists, and the time at which 101 actually exists is It is not the moving speed of the input device. Therefore, it does not make sense to connect all the detected moving speeds of the image input device with a curve as in the conventional method shown in FIG. Further, when the detection speed is sparse, the change in the speed of the curve passing through all the points becomes extremely large as shown in FIG. Actually, the speed change of the image input apparatus hardly becomes drastic, and it is impossible to obtain a clear image even if distortion correction is performed at such a speed.

However, according to the present invention, an average smooth speed change can be detected even for the sparse speed changes shown in FIGS. 201 to 209, and the speed change shown in FIG. 1 is obtained. Hereinafter, the speed change detecting method of the present invention will be described with reference to FIG. The horizontal axis of the graph in FIG. 3 represents time, and the vertical axis represents speed. Normally, when operating the image input device, the speed of the image input device at the start is zero. Therefore, the speed and acceleration of the image input device at the start 300 are set to 0 as initial conditions. Next, start 300 and first detection speed 3
Calculate a quadratic function through 01. Since the quadratic function can be obtained by knowing the position information on the graph of two points and the inclination of one point, that is, the acceleration, it can be found by anyone, so it will not be described here. As a result, a function 341 connecting the position 300 and the position 301 on the graph of FIG. 3A can be obtained.
In the present invention, only the area 311 (half of the position 300 to the position 301) of the function is set as the detection speed, and the areas 311 and 311
The speed and acceleration 321 at the 12th boundary are set as conditions for estimating the following function. Next, when the next detected speed 302 is known, a quadratic function passing through the position 321 and the position 302 is obtained by using the information on the final speed and the acceleration 321 and the information on the speed 302. As a result, a function 342 that passes between the position 321 and the position 302 in FIG. 3B can be obtained. Here, only the portion of the function 342 in the area 315 (half of the position 321 to the position 302) is set as the detection speed, and the speed and acceleration 322 at the boundary between the area 315 and the area 316 are set as conditions for estimating the next function. Here, the function area 311 shown in FIG. 3A and the function area 315 shown in FIG. 3B are connected continuously. This is position 3
Because the function was obtained under the condition of speed and acceleration of 22,
Thereby, a smooth speed change of the actual image input device can be reproduced. Also, by using the obtained function only for half the area and using the speed change of the function obtained based on the next detected speed, the remaining area always focuses on the next speed This makes it possible to reduce the influence of one detection speed, prevent a drastic change in speed, and detect an average speed. Next, when the next detection speed 303 is known, the positions 322 and 30 in FIG.
Find the quadratic function 343 passing through 3. In this case also the quadratic function 343
Only the portion of the region 313 is the moving speed of the image input device,
The speed and acceleration of the position 323 are conditions of a quadratic function when the next speed is known. Similarly, a quadratic function 344 passing through the detected speed 304 and the position 323 is obtained, and the speed is estimated. If N is 1, if N is 1, the first time is the time corresponding to the position 300 on the graph, T (N + 1) at which the speed is detected, that is, T (2 ) Is the position 301 on the graph as shown in FIG.
Is the time corresponding to. The intermediate time, the (N + 1) th, that is, the second time is a time corresponding to the position 321 on the graph. The speed of the image input device is estimated to be a speed indicated by a solid line using a curve approximated to pass through positions 300 and 301 on the graph from positions 300 to 321 on the graph. Then N is 1
If the number is increased to 2, the Nth, that is, the second time is a time corresponding to the position 321 on the graph, so that T (N + 1), that is, T (3) is a time corresponding to the position 302 on the graph. When N is 2, the (N + 1) th is the third, and the time corresponding to the position 322 on the graph is the third time. Hereinafter, similarly, T (4) is the time corresponding to the position 303 on the graph, T (5) is the time corresponding to the position 304 on the graph, and the fourth time corresponds to the position 323 on the graph. The time and the fifth time are times corresponding to the position 324 on the graph. By obtaining a quadratic function each time the speed is known as described above, it becomes possible to finally obtain the speed change shown in FIG. By correcting the input image based on the speed change function, a clear image without distortion can be obtained.

Further, in the present invention, since the function is always obtained using only the information of two points, the amount of the image stored in the buffer can be small even when the distortion is corrected by detecting the speed while inputting the image. Further, since the speed change is estimated from the information of two points, very high-speed calculation is possible. Further, in the present invention, the speed is always estimated by paying attention to the speed ahead, so that a smooth and average change in speed can be estimated. In order to detect the speed originally, the method of detecting the speed at a certain time interval and detecting the average speed is a method of detecting the speed closest to the actual speed change. In the method of detecting the speed by comparing the obtained images, it is impossible to detect the speed at regular intervals. However, when the detection speed is concentrated in a short time by using this method, the effect of each is small when the detection speed is short, and when the detection speed is not detected for a long time, the effect of the detected speed becomes large. Is very close when the average speed is calculated and the speed change is estimated. Therefore, there is no image distortion after correction, and the image becomes very beautiful.

[Effects of the Invention] As described above, according to the present invention, each time a new detected speed is known, a function that passes through the final estimated speed position and the newly detected speed is estimated. Since the function up to the intermediate point between the final estimated speed position and the newly detected speed position is used as the estimated speed, and the speed estimation is repeated using the speed and acceleration of the new final estimated speed position as the conditions of the next function, the actual image It is possible to estimate a continuous speed change very close to the operation speed change of the input device. As a result, it is possible to provide an image without distortion.

According to the image reading method of the present invention, an encoder for positioning or the like becomes unnecessary, so that a small-sized image input device can be further miniaturized and an extremely easy-to-use image input device can be provided.

[Brief description of the drawings]

FIG. 1 is a diagram showing a speed change estimated by the method of the present invention. FIG. 2 is a diagram showing a speed change estimated by a conventional method. FIG. 3 is an explanatory diagram of a speed change estimating method according to the present invention.
FIG. 4 is a configuration diagram of an image reading apparatus used in the present invention. Fifth
The figure is an explanatory view of an image input by the image reading device. 101-109 detection speed 210-209 detection speed 301-304 detection speed 401 sensor interval 402 second sensor 403 first sensor 404 second sensor reading position 405 second 1 sensor reading position 406 …… Lens 407 …… Read original image 501 …… Time axis

Claims (1)

(57) [Claims] An image sensor comprising at least two first and second sensors, each of which is a one-dimensional image sensor for reading characters and graphics, is read at a position where the first sensor moves forward in the manual feeding direction. The image is moved to the second
An image reading method of an image input device that detects a time until a sensor reads, estimates a scanning speed from the time, corrects an input image for distortion, and reads an image, wherein a horizontal axis represents time, and a vertical axis represents speed. When assuming, a function approximating a curve passing through the estimated speed at the N-th time estimated last by the image input device and the detected speed at the time T (N + 1) after the N-th is estimated. An intermediate time between the Nth time and the time of T (N + 1) is set to the Nth time.
As the time of +1, the estimated speed of the image input device is the speed of the curve represented by the function from the Nth time to the N + 1th time, and the N (1, 2, 3,...)
The scanning speed of the image input device is approximated by a curve of the function by sequentially increasing the value of (integer).