Differences between point, line, and kernel density
You can calculate density using either simple or kernel calculations. In a simple density calculation, points or lines that fall within the search area are summed, then divided by the search area size to get each cell's density value.
The difference between the Point Density and Line Density tools is that the first is applied to point features and the second to linear features. The two calculate the quantity specified by the Population field that falls within the identified neighborhood and divide that quantity by the area of the neighborhood.
The difference between the output of those two tools and that of Kernel Density is that in point and line density, a neighborhood is specified that calculates the density of the population around each output cell. Kernel density spreads the known quantity of the population for each point out from the point location. The resulting surfaces surrounding each point in kernel density are based on a quadratic formula with the highest value at the center of the surface (the point location) and tapering to zero at the search radius distance. For each output cell, the total number of the accumulated intersections of the individual spread surfaces is calculated.