Delay-and-sum (DAS) beamforming is unable to identify individual scatterers when their density is so high that their point spread functions overlap each other. This paper proposes a convolutional neural network (CNN)-based method to detect and localize high-density scatterers, some of which are closer than the resolution limit of DAS beamforming. A CNN was designed to take radio frequency channel data and return nonoverlapping Gaussian confidence maps. The scatterer positions were estimated from the confidence maps by identifying local maxima. On simulated test sets, the CNN method with three plane waves achieved a precision of 1.00 and a recall of 0.91. Localization uncertainties after excluding outliers were ±46μm (outlier ratio: 4%) laterally and ±26μm (outlier ratio: 1%) axially. To evaluate the proposed method on measured data, two phantoms containing cavities were 3-D printed and imaged. For phantom study, training data were modified according to the physical properties of the phantoms and a new CNN was trained. On an uniformly spaced scatterer phantom, a precision of 0.98 and a recall of 1.00 were achieved with the localization uncertainties of ±101μm (outlier ratio: 1%) laterally and ±37μm (outlier ratio: 1%) axially. On a randomly spaced scatterer phantom, a precision of 0.59 and a recall of 0.63 were achieved. The localization uncertainties were ±132μm (outlier ratio: 0%) laterally and ±44μm with a bias of 22μm (outlier ratio: 0%) axially. This method can potentially be extended to detect highly concentrated microbubbles in order to shorten data acquisition times of super-resolution ultrasound imaging.