Feature Description
Image stretching is an image/imagery tool that can change the brightness and contrast of images/imagery, thereby improving their quality, making features in the imagery easier to identify, and more conducive to AI target recognition.
The big data component provides four stretching tools: standard deviation, minimum maximum, histogram equalization, and percent clip.
The selection of tools and parameters such as statistical values are obtained based on experience or pre-testing. For example, the adjustment of stretching parameters can be performed in real-time browsing on the desktop client or web client, and after adjusting the parameters, they can be applied to subsequent automated processing workflows. The statistical values required for stretching may come from basic statistics, histogram calculations, directly obtained from the statistical values stored in the mosaic dataset, or manually input by users (understanding that there may be numerical distortions in a few images among multiple datasets, in which case using global statistical data for stretching is not the best choice).
Additionally, image stretching is often used in pixel format conversion, such as stretching higher pixel formats (e.g., 16-bit) to 8-bit, or stretching floating-point types with a range of 0 to 1 to 8-bit. After stretching, it may need to be combined with pixel format conversion.
Image minimum maximum is the linear stretch of the minimum and maximum values. This method uses the minimum and maximum values of the pixel values as the range domain, performs linear stretch, so that the pixel values are distributed between [0,255]. Through such stretching processing, the contrast and brightness of the imagery are significantly improved, making features in the imagery easier to identify. It is generally suitable for stretching raster images with dense pixel value distributions. The returned result type is a raster dataset (RasterRDD).
Linear stretch implementation: Calculate the normalized values and then scale them to the range of the maximum and minimum output values.
Parameter Description
| Parameter Name | Default Value | Parameter Interpretation | Parameter Type |
|---|---|---|---|
| RDD to Analyze | RDD to analyze. | RasterRDD | |
| Whether to Perform Statistics Within the Function (Optional) |
true | Whether to perform statistics within the function, default is true, meaning that full statistics are performed on the input raster dataset within the function. false indicates that statistical data is passed in externally. | Boolean |
| Minimum Value of Each Band (Optional) |
The original minimum value of each band of the data, with the array length equal to the number of bands. Effective only when [Whether to Perform Statistics Within the Function] is not selected. This value can be an empirical value for certain types of data, calculated from basic statistics or histograms, or directly obtained from the statistical values stored in the data information. | Object | |
| Maximum Value of Each Band (Optional) |
The original maximum value of each band of the data, with the array length equal to the number of bands. Effective only when [Whether to Perform Statistics Within the Function] is not selected. This value can be an empirical value for certain types of data, calculated from basic statistics or histograms, or directly obtained from the statistical values stored in the data information. | Object | |
| Minimum Target Pixel Value | 0 | The minimum target pixel value, original pixel values are stretched between the maximum and minimum target pixel values. Default is 0. | Integer |
| Maximum Target Pixel Value | 255 | The maximum target pixel value, original pixel values are stretched between the maximum and minimum target pixel values. Default is 255. | Integer |