摘要:
871,163. Automatic character recognition. STANDARD TELEPHONES & CABLES Ltd. Jan. 2, 1959 [Jan. 9, 1958], No. 182/59. Addition to 871,162. Class 106 (1). In apparatus for automatically reading characters where signal patterns representing the characters are stored in a two-dimensional shift register and are displaced to bring the pattern into registration with a predetermined position for identification as described in the parent Specification, the characters are scanned in columns (or lines) and the signals derived are stored in turn in an end column (or the top or bottom line) of the shift register and shifted into the register as succeeding columns (or lines) of the character are scanned. Series of photo-cells.-In the form of Fig. 1 the record 11 is moved past a sensing station where the characters are illuminated by a lamp 13 and their reflected image is dissected by a column of photo-cells 16. Other forms of sensing may be used. The photo-cell outputs are connected through amplifiers O1-O9, Fig. 2, to the left-hand column of storage elements 51-59 of the shifting register 18. When a character comes into the sensing station a signal appears on one or more of the photo-cells and this is passed through Or gate 19 to a generator G1 and via inverter 20 to And gate 21. On receiving this signal the generator G1 produces x (horizontal) shift pulses to allow the photo-cell signals to pass into the first column of storage elements 51-59. Subsequent shift pulses displace these signals to the second, third, fourth and fifth columns in turn and read the photo-cell signals at spaced instants into the first column. By this means the whole character pattern is stored in the shifting register. When the character has passed the sensing station no photo-cell in column 16 produces a signal; there is no output from gate 19 and generator G1 therefore produces no further x shift pulses. Vertical centring.-The absence of output from gate 19 between characters causes inverter 20 to enable gate 21 controlling the generator G2 of y (vertical) shift pulses to adjust the vertical position of the character pattern. The outputs of each element of the five top lines of the register are connected to five Or gates 22-26 each of which must deliver an output when the character is in the uppermost position touching the top line. The gate outputs are connected to an And gate 27 which gives a signal only when all five lines have signals. The character pattern is displaced upwardly until this is so, the y generator G2 being enabled by a signal from gate 21 when a zero from gate 27 causes inverter 28 to produce an output. When the character pattern is positioned with parts in each of the five top lines, the signal from gate 27 stops the generator. Five lines are used for this purpose instead of merely the top line in case a part of a character in the line above is giving a signal in the uppermost line, which, where only the top line was used would produce an indication that the character was in the required uppermost position. Centring on an unwanted character in the line above or below is prevented by a circuit 29 having a counter 30 which counts the number of vertical steps in each centring operation and stores this number in store 33 until the next. Comparator 31 compares each count with the previous one and permits the generator G2 to stop only if the new number differs by less than a predetermined value from the previous one. Single photo-cell.-In the form of Fig. 3, the record 11 is moved past the sensing station as before but is sensed by a beam of light from C.R.T. 34 moving in vertical scans, reflected light being received in a photo-cell 36. A ring counter 38, Fig. 4, is stepped by one of two generators G3, G4 and the stages are connected through resistors W and diodes D to provide the scanning voltage for C.R.T. 34. The counter stages are also connected to step a switch 37 through a corresponding cycle of connections, the photo-cell signals being applied to the switch and distributed to storage elements 10-90 in turn. The occurrence of a " black " signal in any storage element, indicating that a character has arrived in the sensing station, causes Or gate 39 to deliver a signal to And gate 41 the other input of which comes from the " O " stage of counter 38. The signals arriving in serial form from photo-cell 36 corresponding to sample points on a vertical scan through the character are thereby staticized in store 10-90 and at the end of the scan shifted into the first column of the shifting register 18. The output from Or gate 39 also determines which generator G3 or G4 shall be used to step the counter. While no signals are being produced by the photo-cell, i.e. before the character arrives, the high-speed generator G3 is used to give a very rapid scan. As soon as a signal is given the output from gate 39 switches generator G3 off through inverter 40 and switches the slow-speed generator G4 into operation.
摘要:
845,106. Photo-electric character reading apparatus. LABORATORY FOR ELECTRONICS Inc. July 13, 1956 [July 21, 1955], No. 21824/56. Class 40(3). [Also in Group XIX] In a character recognition apparatus, e.g. for reading numbers on a bank cheque and controlling printing, sorting or tabulating devices, the character is scanned photo-electrically in a series of lines as indicated at 1-12 in Fig. 7 and the resulting long and short "black" pulses counted to establish in each line scan: (1) the total number of pulses both long and short, and (2) the number of long pulses (e.g. for scan 3 the totals are respectively 2 and 0, whilst for scan 8 they are 1 and 1). The results of the line scans are then compared according to a programme (see below) and the results inserted in a special shift-register store (see below) which is advanced only at each entry. At the completion of the scan of a character a pattern of activated stages exists throughout the store, the pattern being different for each character, and is interpreted to produce a potential on the appropriate one of a group of terminals. The interpretation is effected by means of a matrix interrogation circuit, Fig. 9 (not shown), in which groups of rectifiers are connected to the various possible pattern of stages in the store and control the terminals via gate stages. According to one programme, an entry is made in the store only when the result of adjacent scans are different, provision being made to enter the first two scans irrespective of identity so as to allow for possible inaccuracies in the first scan which may not wholly coincide with the character. The store is arranged as illustrated in Fig. 6 in a group of five columns, each of seven stages, and entry is made in the various columns as shown in accordance with a code number indicating the results of the two counts. The first digit of the code number indicates to the total number of pulses and the second digit indicates the number of long pulses. The shift control, which is applied only when an entry is made, influences all stages simultaneously so that the pattern as it is built up travels as a whole, down the store. The result of this programme for the number 3 scanned as shown in Fig. 7 is set out scan-by-scan in Fig. 8. The shift control is applied during blank scans 9-12, no entry being made however and the emergence of the code for the first scan at the bottom of a column (column 1 for the example considered) serves to indicate that the stored pattern is ready for interpretation and causes the matrix interrogation circuit to be brought into operation. Fig. 6 shows the store with the pattern for number 3 at the moment of interrogation. The code for the first scan is not used in the interpretation since it may contain inaccuracies and merely serves as explained above to bring about the interrogation. The programme ensures that the second scan is always entered. The process of counting pulses renders the apparatus insensitive to the vertical height and position of the characters, and process according to the programme of making a store entry only when adjacent scans differ renders the apparatus insensitive to character width and speed of travel relative to the scanning station. According to an alternative programme, there must be two identical scans before an entry is made in the store, subsequent adjacent identical scans not causing an entry. In the case of number 3 scanned as in Fig. 7, scan 3 would be entered (because it is the same as 2), not scans 4 and 5 (because they are the same as 3), and scan 7 (because it is the same as 6). This programme results in less store entries and thus reduces the number of stages required, inherently rejects the first (possibly inaccurate) scan, and by virtue of the integrative effect over two scans is stated to render the apparatus insensitive to the effects of dust particles &c. Fig. 5 (not shown) illustrates, in block diagram form only, an apparatus set up to operate according to the first programme described above. The counting is effected by binary counters, two being provided for each count which operate alternately with one retaining the count of the preceding line to permit comparison. The comparison, coding and various switching operations are effected by conventional rectifier-type gating and buffer stages. The shift-register store comprises a chain of binary stages, preferably saturable magnetic-core elements. The scanning station Fig. 3 (not shown), comprises a vertical line of photo-electric cells under which the medium bearing the characters is traversed, the cells being brought into operation cyclically in rapid succession so as to produce the effect of a line scan. The cells respond only to infra-red so as to avoid the effects of ambient lighting.
摘要:
A method for contactless measurements of a vehicle wheel assembly by acquiring a sequence of images as the vehicle wheel assembly moves within a projected pattern of light. Images of the vehicle wheel assembly are acquired and processed to identify the portions of the images corresponding to the wheel assembly, such as by recognition of the wheel rim edge. The identified portion of each image is cropped and a resulting point cloud of data rotational aligned by an optimization procedure to remove the effect of wheel translation and rotation between each image, as well as to identify a center of rotation and amount of rotation for each image which yields a best-fit result. Superimposing the resulting point clouds produces a generated image with a high density of data points on the optimally fit surfaces of the wheel assembly, which can be used to further refine the axis of rotation determination.
摘要:
An embodiment includes at least one computer readable storage medium comprising instructions that when executed enable a system to: receive (a)(i) first radio signal location data for a first object from a radio sensor; and (a)(ii) first visual signal location data for the first object from a camera sensor; perform feature extraction on (b)(i) the first radio signal location data to determine first extracted radio signal features; and (b)(ii) the first visual signal location data to determine first extracted visual signal features; solve a first association problem between the first extracted radio signal features and the first extracted visual signal features to determine first fused location data; and store the first fused location data in the at least one computer readable storage medium. Other embodiments are described herein.
摘要:
An apparatus and method for providing image primitives, such as edge polarity, edge magnitude, edge orientation, and edge displacement, and derivatives thereof, for an object are described. The data are obtained substantially simultaneously and processed in parallel such that multiple objects can be distinguished from one another in real time.
摘要:
The systems and methods of the present disclosure enable a user to use a mobile device to automatically capture a high resolution image of a rectangular object. The methods include capturing a low resolution image of the rectangular object and detecting edges of the rectangular object in the low resolution image, where the edges form a quadrangle, calculating a coordinate of each corner of the quadrangle, calculating an average coordinate of each corner of the quadrangle in a most recent predetermined number of low resolution images, calculating a dispersion of each corner of the quadrangle in the most recent predetermined number of low resolution images from a corresponding coordinate of each calculated average coordinate, determining whether the dispersion of each corner of the quadrangle is less than a predetermined value, capturing a high resolution image of the rectangular object when it is determined that the dispersion of each corner of the quadrangle is less than the predetermined value, and geometrically transforming the quadrangle of the rectangular object in the high resolution image into a rectangle.
摘要:
Methods and apparatuses for compressive sensing that enable efficient recovery of features in an input signal based on acquiring a few measurements corresponding to the input signal. One method of compressive sensing includes folding an image to generate first and second folds, and recovering a feature of the image based on the first and second folds without reconstructing the image. One example of a compressive sensing apparatus includes a lens, a focal plane array coupled to the lens and configured to generate first and second folds based on the image, and a decoder configured to receive the first and second folds and to recover a feature of the image without reconstructing the image. The feature may be a local geometric feature or a corner. Compressive sensing methods and apparatuses for determining translation and rotation between two images are also disclosed.
摘要:
A system and related method for acquiring high quality images of the iris of an unconstrained subject comprising a camera; a controllable focusing component; a focus controller component that controls the lens to focus at successively different points within a focus range, such focus control performed without any input from measurement of whether the image is in focus or out of focus, be it based from measurements of the image or other distance metrics to the subject; and a sharpness detection component that rejects the most out-of-focus images based on measurement of focus on the image is disclosed.
摘要:
A motion sensing device for sensing infrared rays, the motion sensing device includes a substrate; a sensing unit, configured on the substrate for sensing the infrared rays; a stabilizing layer, covering on the sensing unit for fixing and protecting the sensing unit, wherein the stabilizing layer has an opening; a protection layer, formed on the opening; and a coating layer, covering the stabilizing layer for absorbing infrared rays, wherein the coating layer does not cover the opening.