Claims
- 1. A method for sorting granular materials from a supply in accordance with surface characteristics as measured with a plurality of individual grain observation devices, each of which generates amplitude values related to the intensity of a colour reflected from an individual grain, comprising the steps of:
- passing individual grains from the supply one after another past the grain observation devices to generate from each thereof a plurality of measured amplitude values related to the intensity of a colour reflected from an individual grain;
- generating standard amplitude values for different colours reflected from a representative grain of the supply;
- combining, for different colours reflected from a grain and detected by a device, the plurality of measured amplitude values associated with a column grain and a common colour to form measured mean amplitude values for the respective colours detected by the observation devices;
- determining the differences between said standard amplitude values and corresponding colour related measured mean amplitude values to derive deviations of the measured mean amplitude values relative to said standard values as a function of colour for the grains in the supply;
- adding the mean amplitude values associated with a common grain but for first and second different colours to form a first coordinate value;
- forming the difference between said mean amplitude values associated with the common grain for said first and second different colours to form a second coordinate value; and
- repeating said adding and forming steps for other grains and recording said first and second coordinate values for the grains passed through the observation devices.
- 2. The method for sorting granular materials as claimed in claim 1 wherein the step of generating standard amplitude values comprises the step of:
- combining a plurality of measured amplitude values associated with different colours of the representative grain to form mean standard amplitude values respectively associated therewith; and
- wherein said difference determining step is done between the mean standard amplitude values and corresponding colour related measured mean amplitude values.
- 3. A method for sorting granular materials from a supply in accordance with surface characteristics as measured with a plurality of individual grain observation devices, each of which generates amplitude values related to the intensity of a colour reflected from an individual grain, comprising the steps of:
- passing individual grains from the supply one after another past the grain observation devices to generate from each thereof a plurality of measured amplitude values related to the intensity of a color reflected from an individual grain;
- combining, for different colours reflected from a grain and detected by a device, the plurality of measured amplitude values associated with a common grain and a common colour to form measured mean amplitude values for the respective colours detected by the observation devices;
- repeating said combining steps for each of the different colours detected by said devices;
- adding the mean amplitude values associated with a common grain but for first and second different colours to form a first coordinate value;
- forming the difference between said mean amplitude values associated with the common grain for said first and second different colours to form a second coordinate value; and
- repeating said adding and forming steps for other grains and recording said first and second coordinate values for the grains passed through the observation devices.
- 4. The method as claimed in claim 3 and further comprising:
- accumulating first and second coordinate values associated with different grains to define a population map associated with said grains in the supply.
- 5. An apparatus for sorting granular materials from a supply in accordance with surface characteristics as measured with a plurality of individual grain observation devices, each of which generates amplitude signals related to the intensity of a colour reflected from an individual grain, comprising:
- means for combining the amplitude signals associated with a common grain and a common colour thereof to produce mean values for a plurality of colours as detected by said observation devices for the common grain;
- means for generating a first coordinate signal representative of the difference between mean values associated with the common grain but different colours; and
- means for generating a second coordinate signal representative of the sum of the mean values associated with the common grain and with the same different colours used to generate the first coordinate signal.
- 6. The apparatus of claim 5 and further comprising:
- means for generating standard amplitude signals for different colours reflected from a representative grain of the supply;
- means for removing, from said mean values, the standard amplitude signals of corresponding colours to produce signals representative of the deviations of the amplitude signals relative to respective standard amplitude signals.
- 7. A method of sorting a batch of granular material on the basis of the colorimetric characteristics of the grains of said batch observed through a number z of color bands comprising the steps of:
- separating said grains from each other;
- advancing said grains through an observation area;
- observing each said grain with a plurality of n observation devices during passage through the observation area in which a background is placed opposite each observation device so that the said grain passes between each said observation device and its respective background, each said device generating a number z of color signals obtained from each said observation device on passage of each grain;
- sampling m times and digitizing the color signals to obtain m values from each color signal;
- summing the m values for each color signal to obtain a total number of values equal to the total number z.times.n of generated color signals;
- observing a group of grains that is a representative sample of the batch to generate a mean value of each color signal;
- deducting from each said z.times.n values the mean value of the respective color signa;
- processing the sum of the values corresponding to the same color band to obtain a number z of final color coordinate values which represent the colorimetric characteristics of each grain observed through the single color bands and establish a point representative of the colorimetric characteristics of each grain in a cartesian datum system of z coordinates defining a color space;
- adding two final color coordinate values when the number z of color bands is two to obtain a number A representative thereof; and
- subtracting one final color coordinate value from the other to obtain a number C representative thereof, said numbers A and C representing the colorimetric characteristics of each observed grain and establishing a point on a cartersian datum system having first and second coordinates representative of the colorimetric characteristics of each said grain.
- 8. A method of sorting a batch of granular material on the basis of the colorimetric characteristics of the grains of said batch observed through a number z of color bands comprising the steps of:
- separating said grains from each other;
- advancing said grain through an observation area;
- observing each said grain with a plurality of n observation devices during passage through the observation area in which a background is placed opposite each observation device so that said grain passes between each said observation device and its respective background, each said device generating a number z of color signals obtained from each said observation device on passage of each grain;
- sampling m times and digitizing the color signals to obtain m values from each said color signal;
- summing the m values for each color signal to obtain a total number of values equal to the total number z.times.n of generated color signals;
- observing a group of grains that is a representative sample of the batch to generate a mean value of each color signal;
- deducting from each said z.times.n values the mean value of the respective color signal;
- processing the sum of the values corresponding to the same color band to obtain a number z of final color coordinate values which represent the colorimetic characteristics of each grain observed through the single color bands and establish a point representative of the colorimetric characteristics of each grain in a cartesian datum system of z coordinates defining a color space, an electronic grid composed of a finite number of computer memory cell elements being superimposed on said color space so as to cover the latter completely and in such a way that every point of said color space, as represented by said color coordinates, may be associated with one only of said cells;
- supplying said representative sample in the machine;
- clearing all said cells;
- observing the grains belonging to said sample;
- observing and calculating the color coordinate for each said grain; and
- detecting the cell in said electronic grid which corresponds to said color coordinate and, in such case, increasing by one unit the value contained in that cell so that, upon completion of the observation of said sample, each single cell contains a value equivalent to the number of grains having the colorimetric characteristics corresponding to that cell,
- whereby in the aggregate said cells in the grid provide statistical information in the form of a population map that is indicative of the distribution of said sample in the color space.
- 9. The method according to claim 8 wherein the number z of color bands is two so that the color space has two dimensions, the color coordinates, being represented by numbers A and C to defines a cartesian datum system (A, C) so that, upon completion of the observation of said sample, a matrix comprising a population map of said batch is obtained.
- 10. The method according to claim 9, comprising the steps of:
- forming an electronic grid selection map that is superimposed to cover said color space in the same way the population map does;
- forming each cell in the selection map so as to take only the value zero or the value one;
- classifying an observed grain as acceptable, if the colorimetric characteristics of that grain coincide with a value 0 cell and to be not acceptable if the colorimetric characteristics of that grain coincide with a value 1 cell, said values 0 and 1, depending on a programmed selection standard; and
- causing an unacceptable grain to be expelled by the sorting machine.
- 11. The method according to claim 10 which comprises applying grain selection standards that are independent from one another and include:
- i. expelling the grains which have an A value higher than a certain predetermined value;
- ii. expelling the grains which have an A value lower than a certain predetermined value;
- iii. expelling the grains which have a C value higher than a certain predetermined value;
- iv. expelling the grains which have a C value lower than a certain predetermined value; and further comprising the steps of:
- programming said predetermined values, superimposing the different value 1 cells corresponding to different selection standards, only the cells corresponding to a grain expulsion condition according to the chosen standard being given value 1 automatically in the selection map, the remaining cells being unchanged.
- 12. The method according to claim 10 which comprises the selection standard of: observing in the observation devices a group of grains considered--to be expelled--, giving value 1 all the cells in the selection map which correspond to the colorimetric characteristics of the grains of said group, automatically expelling all the grains in a batch under sorting which have colorimetric characteristics equivalent to the colorimetric characeristics of the grains of said group.
- 13. The method according to claim 10 which comprises the selection standard of defining as value 1 cells in the selection map all the cells corresponding to cells in the population map which contain a value lower than a predetermined value, all those grains in a batch under sorting whose presence in said batch is considered too low being so expelled.
- 14. The method according to claim 10 comprising the steps of: processing the sum of all the cells in the population map which coincide with value 1 cells in the selection map and calculating the equivalent percentage value which provides a statistical forecasting about the number of grains that will be expelled form a batch.
- 15. A method for sorting granular materials by separating the grains one from another; passing the grains along channel means and in a lit observation cell; repeatedly viewing each grain in z color bands in a group of n observation devices within the lit observation cell when said grain passes in the cell; generating a number z of color signals from each said n observation devices on passage of each grain; sampling m times and numerically converting said z color signals from each grain to obtain m values from each said z color signals; the method being characterized in that it comprises the further steps of: reducing the plurality of signals obtained for each grain into z color coordinates containing the color characteristics of each grain; viewing a group of grains that is a representative sample of a grain batch and generating in an electronic computer a population map having z color coordinates and containing the density of probability of each color coordinate in the grain batch; establishing a selection standard and generating a selection map in z dimensions, which selection map covers the color space defined by said z dimensions, each cell in the selection map taking only the value 0 or the value 1, an observed grain being classified as--acceptable--if its color coordinates coincide with a value 0 cell, an observed grain being classified as--to be expelled--if its color coordinates coincide with a value 1 cell and being expelled by an ejector device in the machine.
- 16. Method according to claim 15 characterized in that an electronic grid composed by a number of memory cell elements is set in the computer and is superimposed to said color space in such a way that every point of said color space may be associated with one only of said cells, the method comprising the further steps of: after viewing the grains belong to said representative sample, calculating the color coordinate of each viewed grain, detecting that cell in said electronic grid which corresponds to said color coordinate, increasing by one unit the value contained in that cell so that, on completion of viewing said representative sample, each single cell contains a value equivalent to the number of grains having the color characteristics corresponding to that cell, the group of said cells supplying a statistical information about the spreading of said representative sample in said color space.
- 17. The method according to claim 15 characterized in that said color coordinates, as obtained under the condition that the number z of color signals is 2, are two values which are added one to the other to obtain a number called A and are subtracted one from the other to obtain a number called C, said numbers A, C representing the color characteristics of each viewed grain and esablishing a point in a map having first and second coordinates which are representative of the color characteristics of each said grain.
- 18. The method according to claim 17 characterized in that, said color space having two dimensions and the color coordinates being represented by said numbers A, C to define a cartersian datum system (A, C), a population map of said grain batch is set up in the memory of the computer.
- 19. The method according to claim 15 characterized in selection standards that are applicable of:
- expelling the grains which have a A value higher than a certain predetermined value,
- expelling the grains which have a A value lower than a certain predetermined value,
- expelling the grains which have a C value higher than a certain predetermined value,
- expelling the grains which have a C value lower than a certain predetermined value,
- programming said predetermined values, combining at will part or all the above selection standards in order to superimpose the different value 1 cells corresponding to different combined selection standards, only the cells corresponding to a grain expulsion condition according to the chosen standard being given value 1 automatically in the selection map, the remaining cells being unchanged.
- 20. The method according to claim 15 characterized in that the selection standard is applicable of: viewing a group of grains considered--to be expelled--, giving value 1 all the cells in the selection map which correspond to the color characteristics of the grains of said group, automatically expelling all the grains in a batch under sorting which have color characteristics equivalent to the color characteristics of the grains of said group.
- 21. The method according to claim 15 characterized in that the selection standard is applicable of: defining as value 1 cells in the selection map all the cells corresponding to those cells in the population map which contain a value lower than a predetermined value, all those grains in a batch under sorting whose presence in the batch is considered too low being so expelled.
- 22. The method according to claim 15 characterized in that the steps are comprised of: processing the sum of the values of all the cells in the population map which coincide with value 1 cells in the selection map and calculating the equivalent percentage value which provides a statistical forecast about the number of grains that will be expelled from a grain batch.
Priority Claims (1)
Number |
Date |
Country |
Kind |
24875 A/82 |
Dec 1982 |
ITX |
|
Parent Case Info
This application is a continuation of application Ser. No. 560,821, filed Dec. 13, 1983, now abandoned.
US Referenced Citations (6)
Foreign Referenced Citations (3)
Number |
Date |
Country |
0025284 |
Mar 1981 |
EPX |
885287 |
Dec 1961 |
GBX |
2068535 |
Aug 1981 |
GBX |
Continuations (1)
|
Number |
Date |
Country |
Parent |
560821 |
Dec 1983 |
|