A Thematic Mapper (TM) is one of the Earth observing sensors introduced in the Landsat program. The first was placed aboard Landsat 4 (decommissioned in 2001), and another was operational aboard Landsat 5 up to 2012.  TM sensors feature seven bands of image data (three in visible wavelengths, four in infrared) most of which have 30 meter spatial resolution. TM is a whisk broom scanner which takes multi-spectral images across its ground track. It does not directly produce a thematic map.
The upper photo on the right is a 50 times magnification of the combined photomasks used to fabricate the Hughes H4040, the linear Silicon photodiode array used in the Thematic Mapper to image the visible bands. Each of the 16 photodiodes is 100 microns square and their separation is 100 microns. There are two rows because it is scanned perpendicular to the lines of diodes and they produce a complete line with no separation. The alignment marks and their layer names can be seen at each end. Each layer is a different color. The pink layer is a second layer of Aluminum acting as an aperture. The openings had to be 100u square exactly. The exact dimensions were required in order to achieve a 30 meter resolution on the ground. I Dominic Massetti fabricated a set of four of these and documented the fabrication process to NASA requirements and verified the dimensions as part of my job at Hughes Aircraft Company's Industrial Products Division in Carlsbad California in 1978. The challenge was to customize each of these for one of the narrow visible bands that were required. To do that the thickness of the Silicon Nitride antireflective layer had to be at a precise target and all the photodiodes in the array had to have the same thickness. At the time all we had was an atmospheric deposition system that basically burned Silane (SiH4) in the presence of Ammonia in a horizontal tube heated to about 850 degrees Centigrade. But that process yielded a smokey film that varied over the Silicon wafer and the diode array. So I took a cue from my workmate Paul Hilton who had invented low pressure (LPCVD) polysilicon deposition at Motorola a few years earlier and we built a low pressure Silicon Nitride system using Silane and Ammonia to produce the precisely tunable and uniform thicknesses needed at each of the bands. Without that innovation there would have been no TM. I also had to get the Boron diffused photodiodes to have very low dark current and high photosensitivity in order to meet the imager specifications. The large crosses were used when the different arrays were aligned together at final assembly in the Thematic Mapper. The final assembly or the TM Focal Plane without filers is shown in the second photo to the right.  They flew in Landsat 4 and for 20 years they flew high.
Also shown on the right is a letter of congratulations from Hughes Aircraft Company Industrial Products Division (IPD) on the invention of the LPCVD Silicon Nitride system that made the Thematic Mapper diode arrays possible.
Additionally a photo shows the Hughes (IPD) Newsletter from August 1978 highlighting the second level of Aluminum that formed the light shield and set the aperture locations. It was an accomplishment to prevent this shield from shorting out the lower layer Aluminum leads to the photodiodes through defects and pinholes given that this layer was a large sheet over everything.
The Thematic Mapper has become a useful tool in the study of albedo and its relationship to global warming and climate change. The TM on the Landsat 5 has proven useful in determining the amount of ice loss on glaciers due to melting.
2. J. L. Engle and O. Weinstein, "The Thematic Mapper---An Overview," in IEEE Transactions on Geoscience and Remote Sensing, vol. GE-21, no. 3, pp. 258-265, July 1983, doi: 10.1109/TGRS.1983.350551.