The simple difference between scientific discovery and scientific observation is a matter of qualification, quantification and verification. While observation may be a contributing factor to discovery, the two are otherwise disassociated and even somewhat mutually exclusive.
Anyone can make a scientific observation. An observation by itself is representative only of the perceptions derived from it. It is common for humans to perceive something and then speculate based on their perceptions, but this often leads to erroneous conclusions. For instance, for thousands of years humans perceived the Earth to be flat. It was a valid perception based on valid observations, because on the surface of the Earth our perspective is limited and the curvature of the Earth's surface is undetectable to the eye and unperceived by the brain.
In the first century BCE, the Greek astronomer Aristarchus observed, that during the autumnal equinox, a stick stuck in the ground and pointing straight up in the air cast a longer shadow in his home in Samos than in the city of Alexandria Egypt, where it cast almost no shadow at all. This basically represents at least two instances of observation, neither of which is representative of a scientific discovery. They are simple observations. But when Aristarchus tried to understand the phenomenon better he applied a little geometry and made a major scientific discovery. He new the distance between Samos and Alexandria and hypothesized that if the Earths surface was curved it would explain the difference. In the process of his quantitative analysis, Aristarchus also figured out the circumference and diameter of the Earth and by calculating the parallax, he also deduced the distance between the Earth and Sun. If Aristarchus's observations of shadows cast by sticks were not themselves discoveries, they certainly led to a few profound ones. Unfortunately, Aristarchus's colleagues could not understand what Aristarchus did, and it would take about another 2,000 years before the flat Earth doctrine was revised to conform to his discoveries.
The process that Aristarchus used to make his discoveries may have begun with observations, but until he qualified, quantified and verified his perceptions, the discovery could not be validated. Furthermore, a scientific discovery must be an instance of novelty, a first of a kind perception and it doesn't necessarily have to involve any observations at all. An example would be the discovery of the Big Bang. In the nineteen twenties, the brilliant physicist Monsignor Georges Lamaitre concluded, based on his own understanding of Einstein's theories of special and general relativity, that the universe must be expanding and that that expansion began with a transformation of energy into matter, an event he referred to as a primeval atom. A few years later, Edwin Hubble made some observations with his telescope which gave credence to Lemaitre's primeval atom and expanding universe theory. Based on his many observations, Hubble also made a few discoveries of his own.
Of Course, it is certainly possible that a scientific discovery could be based on a single observation, but in science some method of proof or independent confirmation would be required before it would be recognized as a discovery. Therefore, while a scientific observation can certainly be the first step in the process of discovery, it can never be the only step.