Cognitive load reducing platform having image edge enhancement

A system, comprising:

• a memory containing a thermal image from a thermal camera;
• a processor coupled to the memory, wherein the thermal camera and the processor are worn by a user;
• and
• a software component executed by the processor that is configured to perform:
• receiving a thermal image and generating a gradient magnitude image comprising a plurality of pixels having associated gradient magnitude values;
• partitioning the gradient magnitude image into subregions;
• calculating gradient magnitude statistics for each of the subregions;
• calculating mapping parameters for each of the subregions to equalize and smooth a dynamic range of the corresponding gradient magnitude statistics across the subregions, and interpolating the mapping parameters as a function of computed distances to neighboring subregions to generate smoothed mapping parameters to compensate for arbitrary subregion boundaries, wherein interpolating the mapping parameters comprises:
• defining a set of anchor points (wi) to be applied to all of the subregions;
• for each pixel (p) in a respective subregion, calculating weights for each anchor point by calculating the distance between a current pixel and each of the anchor points;
• and
• calculating the smoothed mapping parameters α=intercept and β=slope for each pixel (p) in the respective subregion as a weighted sum of for the anchor points wi:
• αp=Σi=08f(wi)*αi and βp=Σi=08f(wi)*βi;
• applying the mapping parameters calculated for each of the subregions to pixels in the corresponding subregions to generate enhanced gradient magnitude values having equalized luminosity and contrast, and using the enhanced gradient magnitude values to form a wireframe image having enhanced edges of objects;
• and
• displaying the wireframe image on a display device to the user, wherein the wireframe image appears as a decluttered line drawing where the enhanced edges have increased luminosity and contrast compared to the thermal image to reduce the cognitive load of the user.

A non-transitory computer-readable medium containing program instructions for enhancing edges of objects in a thermal image, wherein the computer readable medium, a thermal camera that captures the thermal image, and a processor for processing the thermal image into a wireframe image in real-time, are worn by a user, the program instructions when executed by the processor for:

• receiving the thermal image and generating a gradient magnitude image comprising a plurality of pixels having associated gradient magnitude values;
• partitioning the gradient magnitude image into subregions;
• calculating gradient magnitude statistics for each of the subregions;
• calculating mapping parameters for each of the subregions to equalize and smooth a dynamic range of the corresponding gradient magnitude statistics across the subregions, and interpolating the mapping parameters as a function of computed distances to neighboring subregions to generate smoothed mapping parameters to compensate for arbitrary subregion boundaries, wherein interpolating the mapping parameters comprises:
• defining a set of anchor points (wi) to be applied to all of the subregions;
• for each pixel (p) in a respective subregion, calculating weights for each anchor point by calculating the distance between a current pixel and each of the anchor points;
• and
• calculating the smoothed mapping parameters α=intercept and β=slope for each pixel (p) in the respective subregion as a weighted sum of α and β for the anchor points wi:
• αp=Σi=08f(wi)*αi and βp=Σi=08f(wi)*βi;
• applying the mapping parameters calculated for each of the subregions to pixels in the corresponding subregions to generate enhanced gradient magnitude values having equalized luminosity and contrast, and using the enhanced gradient magnitude values to form a wireframe image having enhanced edges of objects;
• and
• displaying the wireframe image on a display device to the user, wherein the wireframe image appears as a decluttered line drawing where the enhanced edges have increased luminosity and contrast compared to the thermal image to reduce the cognitive load of the user.

A method for enhancing edges of objects in a thermal image to improve user vision and navigation in obscure environments, comprising:

• computing a gradient magnitude image from the thermal image, the gradient magnitude image comprising a plurality of gradient magnitude values;
• partitioning the gradient magnitude image into a plurality of non-overlapping subregions;
• for each of the subregions, calculating gradient magnitude statistics from the gradient magnitude values in the respective subregion;
• applying predetermined upper and lower bound gradient percentiles and an absolute minimum percentile for the gradient magnitude values across all the subregions;
• for each of the subregions, determining upper and lower gradient magnitude values at the upper and lower bound gradient percentiles, respectively;
• for each of the subregions, calculating mapping parameters to apply to the pixels in the respective subregion to linearly map the upper and lower gradient magnitude values to a predetermined reference gradient magnitude representing a desired dynamic range applied to all the subregions;
• for each of the pixels in each of the subregions, computing a distance to neighboring subregions and interpolating the mapping parameters generated for the respective subregion as a function of the computed distances to generate smoothed mapping parameters;
• applying the smoothed mapping parameters to the pixels in the gradient magnitude image to generate a wireframe image;
• and
• displaying the wireframe images in a line of sight of an eye of a user, wherein the wireframe image appears as a decluttered line drawing where edges have increased luminosity and contrast compared to the thermal image.