Crop particle discrimination methods and apparatus

1. A method of discriminating between kernel and chaff particles, in crop, using a crop viewing apparatus that includes:

a) an image-capturing device that is capable of capturing one or more images, each comprising a plurality of pixels, of crop in the crop viewing apparatus, the image capturing device being operatively connected to an image processing device for processing one or more resulting images;

b) one or more lamps for illuminating crop in the crop viewing apparatus, the or each said one lamp being capable of illuminating the crop in a range of illumination levels up to a maximal level equal to or above a reflectivity saturation level; and

c) one or more control devices for controlling the level of illumination provided by each said lamp, the method comprising the steps of

d) operating at least one said control device to cause illumination of crop in the crop viewing apparatus at one or more level less than the said reflectivity saturation level so as to enhance differences in reflectivity between kernels and chaff in the crop viewing apparatus, and before operating at least one said control device, establishing the or each said one illumination level less than the saturation level, wherein the step of establishing the or each illumination level less than the saturation level includes:

determining for a plurality of images of crop or for a plurality of combinations of images of crop a signal-to-noise (SNR) ratio calculated as the absolute value of the difference between the mean reflectivity of chaff and the mean reflectivity of kernels, divided by the sum of the standard deviations of the reflectivities of chaff and kernels; and

selecting as the illumination level less than the saturation level the illumination level used in the image or in the combination of images that exhibits the highest SNR value;

e) operating the image-capturing device while the crop is so illuminated to capture one or more crop images; and

f) operating one or more said image processing devices in order to discriminate in the image between kernel and chaff particles based on their reflectivity characteristics at a said illumination level less than the saturation level.

2. A method according to claim 1 wherein the at least one processing device establishes the reflectivity as the ratio of the reflection intensity by the kernels or chaff to the reflection intensity by a reference strip.

3. A method according to claim 1 wherein the at least one processing device establishes the reflectivity as the ratio of the reflection intensity by the kernels or chaff to the mean reflectivity of the illuminated crop in the entire captured image.

4. A method according to claim 1 further comprising:

operating the at least one said control device to cause illumination of crop at least one further distinct wavelength and/or illumination level less than the said reflectivity saturation level; and

operating the image-capturing device while the crop is so illuminated to capture at least one further crop image;

wherein the step of operating one or more image processing devices comprises:

operating one or more image processing devices to combine at least two of the captured images in order to maximize differences between kernel and chaff particles in the combination image.

5. A method according to claim 4 , wherein the image processing device or devices make for each pixel a linear combination of the reflectivities in at least two of the captured images.

6. A method according to claim 1 wherein the step of establishing the one or each illumination level less than the saturation level includes:

recording reflectivity characteristics of kernel and chaff particles;

correcting the said reflectivity characteristics for spatial inhomogeneity of illumination at each of a series of illumination levels; and

selecting the illumination level or the combination of illumination levels of the series providing the most pronounced difference in reflectivity between kernel and chaff of a crop type following such correction.

7. A method according to claim 1 , wherein the step of establishing the one or each illumination level less than the saturation level includes:

recording reflectivity characteristics of kernel and chaff particles at a plurality of illumination levels at a plurality of wavelengths;

correcting the said reflectivity characteristics for spatial inhomogeneity of illumination;

generating combinations of the corrected reflectivity characteristics at said plurality of illumination levels and of wavelengths, and

selecting for each wavelength the illumination level of the series used in the generated combination providing the most pronounced difference in reflectivity between kernel and chaff of a crop type.

8. A method according to claim 6 wherein the step of correcting the reflectivity characteristics includes at each of the series of illumination levels determining the ratio of absolute reflection intensity of one or more individual pixels to a mean absolute reflection intensity of an entire image.

9. A method according to claim 1 including the step of using a genetic algorithm to select one or more said illumination levels less than the saturation level.

10. A method according claim 1 including the further steps of:

generating a control command in dependence on the outcome of the method; and

transmitting the control command to an adjustable component or sub-system of a harvesting machine in which the viewing apparatus is included or to which the viewing apparatus is operatively connected in order to alter the setting of the adjustable component or sub-system.